REA Group Ltd v Real Estate 1 Ltd

Date of hearing: 11-13, 16 & 18 April 2012

Counsel for the Applicants: Mr W Houghton QC with Mr S Rebikoff

Solicitor for the Applicants: King & Wood Mallesons

Counsel for the Respondents: Mr G Clarke SC with Ms H Tiplady

Solicitor for the Respondents: J A Lawyers Pty Ltd

IN THE FEDERAL COURT OF AUSTRALIA VICTORIA DISTRICT REGISTRY GENERAL DIVISION VID 900 of 2010

JUDGE: BROMBERG J DATE OF ORDER: 7 JUNE 2013 WHERE MADE: MELBOURNE

1. On or before 21 June 2013, the parties confer and bring in short minutes of orders dealing with: (i) the exchange and filing of submissions as to the extent and form of any relief which ought be granted consequent upon the findings made in the reasons for judgment accompanying these orders; and (ii) the steps to be taken to facilitate any further hearing as to damages. Note: Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011.

IN THE FEDERAL COURT OF AUSTRALIA VICTORIA DISTRICT REGISTRY GENERAL DIVISION VID 900 of 2010

JUDGE: BROMBERG J DATE: 7 JUNE 2013 PLACE: MELBOURNE

REASONS FOR JUDGMENT 1 Scouring the real estate classifieds of a metropolitan daily newspaper used to be an essential Saturday morning activity for those interested in buying or renting real estate. With the advent and increasing popularity of the internet, consumer behaviour has been greatly altered and perhaps no more so than in relation to the purchase or rental of real estate. Over the last ten years the advertising of property has increasingly shifted away from newspaper classifieds to what are known as internet property portals. A property portal is a website where real estate agents list and consumers view residential or commercial property. A property portal enables a user to search for properties by reference to criteria such as location, features and price. 2 The second applicant is a wholly owned subsidiary of the first applicant. The applicants (collectively "REA") have a business which operates two property portals of significance to this case. The first is located on the internet at www.realestate.com.au. That property portal has operated since 1998 and lists residential properties that are available to rent or buy. Both the domain name (a part of the internet address) and the trading name used by REA for this business is "realestate.com.au". That name is the subject of three composite trade marks held by REA. The second property portal, operational since 2002, is located on the internet at www.realcommercial.com.au. This portal lists commercial properties available for purchase or rent. Both the domain name and the trading name used by REA for this part of its business is "realcommercial.com.au". That name is the subject of a composite trade mark held by REA. 3 The first and second respondents also carry on a business which operates property portals. The non-corporate respondents are either principals of the business conducted by the corporate respondents or work in that business. Since March 2009, the corporate respondents (collectively "Real Estate 1") have operated a residential property portal located on the internet at www.realestate1.com.au. The internet address of the business includes the domain name "realestate1.com.au". The business has traded under the name "Real Estate 1" and also by reference to its domain name. 4 Between August 2009 and the commencement of this proceeding in October 2010, Real Estate 1 also operated a property portal dealing with commercial property that was located on the internet at www.realcommercial1.com.au. The internet address for that business included the domain name "realcommercial1.com.au" which it also used as its trading name. Since October 2010 it has instead used "realestate1commercial.com.au" as its domain and trading names. 5 The following major questions arise for determination: (a) Did the use by Real Estate 1 of the name "realestate1.com.au" in relation to real estate advertising services, give rise to conduct that did or that was likely to mislead or deceive internet users searching for REA's residential property portal to regard Real Estate 1's portal at www.realestate1.com.au to be the same, or related to or associated with REA's portal located at www.realestate.com.au, in contravention of (the former) ss 52, 53(c) and 53(d) of the Trade Practices Act 1974 (Cth) ("the TPA")? (b) Did the use by Real Estate 1 of the name "realcommercial1.com.au" in relation to real estate advertising services, give rise to conduct that did or was likely to mislead or deceive internet users searching for REA's commercial property portal to regard Real Estate 1's portal at www.realcommercial1.com.au to be the same as, related to, or associated with REA's portal located at www.realcommercial.com.au, in contravention of ss 52, 53(c) and 53(d) of the TPA? (c) In the circumstances described, did the use by Real Estate 1 of the names "realestate1.com.au" and "realcommercial1.com.au" in relation to their property portals amount to passing off at common law? (d) Did the use by Real Estate 1 (whether alone or as part of a logo) of the marks "realestate1.com.au" or "realcommercial1.com.au" infringe REA's realestate.com.au trade marks or the realcommercial.com.au trade mark contrary to s 120 of Trade Marks Act 1995 (Cth) ("the Trade Marks Act")? (e) Are the individual respondents liable as accessories for any contravention of the TPA, passing off and/or trade mark infringement committed by Real Estate 1?

The internet and other technical background facts 6 The determination of the central issues in this case requires a basic understanding of how the internet, and search engines used to find information on the internet, operate. The evidence dealing with those background matters was not controversial. The following findings are largely drawn from the evidence provided by Michael Hudson, a member of REA's 'Web Strategy' team with extensive experience in website design and development and search engine optimisation. For ease of reference, technical terms are shown in bold. 7 The internet is a system of interconnected computer networks. The series of interlinked documents on the internet are known as the world wide web. A web page is a document containing text, images or other information that is available to be viewed on the world wide web. Each web page has a unique address or URL (uniform resource locator) which is also know as a web address. When a series of related web pages are linked together and designed to be read as an integrated whole, they are known as a website. 8 Internet users who wish to make information available to others on the world wide web can register a domain name, which is an alphanumeric alias for a web address. Most web addresses contain a domain name alongside a component which is used to designate a particular type of server, such as "www" for the world wide web or "m" for a mobile server. 9 Once a domain name is registered, information in the form of a web page or website may be located at that domain name and internet users are able to retrieve information from that page or site by entering the relevant domain address (either with or without the "www" prefix) into a web browser. A web browser is a computer program used for retrieving and displaying information on the internet. Some well known examples of web browsers are Internet Explorer (produced by Microsoft), Firefox (by Mozilla), Chrome (by Google) and Safari (by Apple). 10 A domain name consists of at least two parts, known as labels or levels, which are separated by full stops. The two essential components of a domain name are a top-level domain (for instance ".com", ".org" or ".net") and a second-level (or lower level) domain, which is the specific or unique component of the domain name, selected by the website owner. Typically, the second-level domain will consist of the website owner's trading name or brand name (such as "amazon" in the domain name "amazon.com"). 11 There are some 22 generic top-level domains maintained by the Assigned Numbers Authority, an entity responsible for overseeing global internet address allocation. The top-level domain ".com" is the most popular generic top-level domain on the internet. It is commonly utilised by commercial entities, whilst ".gov" is used for government entities and ".edu" for educational institutions. Many domain names also have a country code top-level domain. The country code top-level domain for Australia is ".au". 12 An internet user searching for a website who is familiar with the specific URL or domain name associated with the website can find the website by entering that URL or domain name directly into the web browser on his or her computer. However, a person unfamiliar with the URL or domain name of a website will ordinarily use a search engine to discover the web address where the information they require may be found. Persons familiar with the URL, may also, for reasons of convenience or habit, use a search engine to find a hyperlink to a website instead of entering the URL directly into a web browser. 13 Search engines are information retrieval systems that index web pages available on the internet using software which enables web pages to be scanned so that the search engine may recall parcels of information associated with those web pages in response to a keyword search request made by a searcher. Well known search engines include Google, Bing and Yahoo!. A search engine searches for information on the internet in response to keywords entered by the user into the search bar of the search engine. Search engines use complex algorithms to find information on the internet relevant to the keywords selected by the user. After conducting a search, search engines display search results on the computer screen of the user in the form of a list. The list may be short or may extend to tens of pages. Each page will generally show a maximum of 10 search results. 14 There are two different types of search results usually displayed on a search results page. Organic search results are search engine results displayed in response to the keyword search terms entered by the user. They are ranked in order of relevance according to a formula applied by the search engine by which the search engine tries to identify the most relevant web pages or websites referrable to the keywords entered by the searcher. Organic search results produced by internet search engines and displayed on the search results page will, for each result shown, usually consist of at least three elements: (a) a headline or title, that when clicked, is a hyperlink to the web page (usually in blue and usually bolded, underlined and in larger type face than the rest of the search result); (b) the URL (including the domain name) for the web page (usually in green and in small type face); and (c) a description of the information on the web page or associated website (usually in black). The following provides an example: Federal Court of Australia www.fedcourt.gov.au/ The Court is a superior court of record and a court of law and equity. 15 Paid search results (also known as sponsored links) are small advertisements that also appear on the search results page partly in response to the specific keyword search terms entered by the user into the search engine. Paid search results are usually presented directly above or on the right hand side of the organic search results but may be presented at the foot of the first page or on the second page. They are usually shaded or separated from the organic search results in order to distinguish them. In addition, the paid search results usually appear under a heading such as "sponsored links", "promotional links" or "ads". Each paid search result is also shown with the same three elements (headline, URL and advertising text or description) as appear in organic search results. 16 If the user clicks on to the headline or title of a search result, the user will be taken to the URL or web address shown in the search result. The web page may or may not be the first page or 'home page' of the website. The search result may provide a link to a different page on a website. For instance a search for "real estate Brunswick" may produce a result that will take the searcher directly to a page on a property portal which lists real estate in Brunswick. In that case the web address given will likely be far longer than the domain name for the website. For instance "www.realestate.com.au/Buy/in-brunswick". URL's appearing in search results will thus vary in size. 17 A further feature of a search results page is that each of the keywords used by the searcher will be highlighted in bold wherever appearing (ie in the title, URL or description). 18 Annexure A to these reasons for judgment is a screenshot of a search results page generated by Google's search engine in response to a keyword search for "Nike shoes". It shows each of the features earlier discussed including the bolding of the keywords. Borders have been added to the screenshot to clearly distinguish between organic and paid results. 19 The mechanisms by which a search engine determines the order in which organic search results and paid search results appear on search results pages require more detailed explanation. 20 Search engines utilise complex algorithms to match information on the world wide web with the keywords selected by the user. There are three main factors that are commonly understood to be taken into account by a search engine in determining the web pages to be included as a search result and the order in which the search results are displayed. Those factors are known as 'trust', 'authority' and 'relevance'. 'Trust' is a measure of credibility, with government websites (which end in the top level domain ".gov") and educational websites (which end in the top level domain ".edu") considered particularly credible because of the nature of the institutions which operate them. 'Authority' is measured by reference to the number of links that a website has from other third party websites, the manner in which those websites link to each other (for example what words are used in the link) and the quality of the content of the linked websites. Thus a website to which other large, credible websites link, is likely to rank more highly than a website without those connections. The 'relevance' of a website is assessed by the search engine considering whether and to what extent the keyword search entered by the user is related to a specific topic, subject or theme covered by the web page or website being considered for inclusion in the search results page. In determining relevance, the search engine will read both the text and images visible to a user looking at a web page and also the underlying source code of the web page. That underlying source code is known as metadata. Metadata may also contain metatags, which are keywords (not visible on the web page) inserted by programmers for the purpose of search engine optimisation, a process I will explain shortly. Search engines also take into account the domain name of the website and information about the person conducting the search, such as location (ascertained from the unique address of the user's computer) and that computer's search history with the search engine. 21 Research suggests that approximately 90% of internet users choose a link that is listed on the first page of the search results screen. An online business therefore has a strong incentive to try to optimise the ranking of its website in the organic search results shown on search results pages. For that reason, website owners will commonly seek to manipulate that ranking through search engine optimisation. This is a process of manually altering a website, or seeking to influence the number and nature of links to the website, so that the website will respond favourably to the algorithms applied by search engines and rank as highly as possible in the list of organic search results. 22 In addition to search engine optimisation, online businesses will commonly purchase paid search results in a process known as search engine marketing. That process is perhaps most usefully explained by reference to the advertising program used by Google known as AdWords. Research shows that approximately 94% of all search engine searches in Australia are carried out using the Google search engine. 23 Google allows advertisers to access the AdWords program to manipulate and monitor the performance of their advertisements. The program uses the pay per click advertising model in which the advertiser is only charged when a sponsored link is chosen or 'clicked' by an internet user. The advertiser selects in advance particular keywords that will trigger a paid search result on a search results page in response to particular keyword search terms entered by a user of Google's search engine. In relation to the words selected, the advertiser indicates the amount that the advertiser is willing to spend if the triggered paid search result is clicked on by the user. This is known as the advertiser's bid. The higher the bid, the more likely it is that the advertisement will appear in the search results when the internet user searches for that keyword or combination of keywords. In circumstances where many advertisers are utilising AdWords in relation to the same keywords, a 'behind the scenes' auction takes place that determines which advertisement will appear in the search results page in response to the user's search request and the order in which those advertisements appear. In addition to the value of the advertiser's bid, other matters are taken into account by Google's paid search algorithm, including the relevance of the advertiser's webpage to the keywords entered by the user. 24 An advertiser may elect to trigger advertisements by exact match, phrase match or broad match. An exact match will trigger a sponsored link only if the keyword search term entered by the searcher is the same as the keyword or keywords selected by the advertiser. A phrase match will trigger a sponsored link if the keyword search includes the keywords selected by the advertiser. For example, if an advertiser selects the keywords "houses for sale", a phrase match may be triggered by the search terms "Edwardian houses for sale" or "houses in Melbourne for sale". A broad match will trigger a sponsored link based on associations between the keywords entered by the user and the keywords selected by the advertiser, as determined by Google's algorithm. For example, if the advertiser's keywords are "houses for sale", a broad match may be triggered by the search terms "homes for sale" or "houses to buy".

REA's property portal businesses - their branding, trade marks and promotion 25 REA's property portal business commenced operation in 1998. Shortly thereafter the business changed the web address of its residential property portal to www.realestate.com.au. At that time the business was small, employing some 8 employees. Only about 40 of the 8,500 residential real estate agents in Australia listed their properties with REA at that time. 26 By 1999, REA had listed on the Australian Stock Exchange and had 34 employees in Sydney, Melbourne and Brisbane. Over the following decade, REA's residential property portal business grew exponentially. By the end of 2002, it was the market leader in the Australian online real estate advertising market. Around that time, one in four real estate agents subscribed to the realestate.com.au property portal. By the end of the 2005 financial year, approximately 80% of all Australian real estate agents subscribed. By 2011, the proportion of real estate agents subscribing to REA's property portals had increased to approximately 95%. 27 Online traffic to www.realestate.com.au also increased exponentially over the same period. In June 2000, the website had approximately 150,000 users. In the month of June 2011, the realestate.com.au website had 6.7 million visits from unique browsers (ie 6.7 million different computers accessed the website). To give some context to the magnitude of use involved, this number exceeded the number of visits made by unique browsers in June 2011 to the websites of some of the most well known daily metropolitan newspapers in Australia, such as The Age and The Australian. In the month of October 2011, the realestate.com.au website had visits from 6.8 million unique browsers and a total of over 16 million visits. 28 Using methodology similar to that used to estimate television ratings, statistical information presented by REA shows that in October 2011, the realestate.com.au website was visited by nearly twice as many people as its closest competitor and by more than six times as many people as its next closest competitor. 29 Until about 2002, commercial property listings were intermingled with residential listings on "realestate.com.au". In 2002, REA launched the realcommercial.com.au portal promoting that business to medium sized commercial property agents and residential agents with some commercial listings. At that time, there were two large and successful dedicated commercial property portals. By 2005, the realcommercial.com.au business was the most popular commercial property portal in Australia. In May 2006, REA acquired one of the other popular commercial real estate portals to consolidate its position as the leading commercial property portal in Australia. In the month of January 2005 some 72,000 unique browsers visited realcommercial.com.au, rising to nearly 400,000 visitors in the month of October 2011. In October 2011, visits to the realcommercial.com.au website were nearly four times the visits to the closest rival of REA's commercial real estate portal. At that time, there were some 1500 subscribing agents to the realcommercial.com.au property portal. 30 In June 2007, REA launched a holiday rental website known as "realholiday.com.au". That website was sold in April 2011 and renamed. REA also operates a website which lists businesses for sale known as "realbusinesses.com.au". 31 REA's various portals refer to each other and contain hyperlinks so that an internet user may conveniently move from one site to the other. 32 There can be, and there was no issue, that REA's residential and commercial real estate portals are popular sites with consumers and are extensively used by consumers to search for properties. 33 Given the popularity and market dominance of REA's real estate portals, its branding has become relatively well known. The branding of a product facilitates a reputational connection between the consumer or potential consumer of a product and the product's producer or provider. Brand recognition by consumers demonstrates an association in the minds of the consumer between the product and its provider or at least between the product and its brand. Some brands will focus on the identity of the provider others will not. Whether focused upon the owner's name or the owner's trading name for the product concerned, a brand enables an owner to establish a relationship between the consumer and the owner or between the consumer and the owner's product or perhaps both. 34 A brand may be comprised of a number of brand elements such as a name, a logo and the use of a particular colour. In the case of REA's residential property portal, there are four main brand elements utilised: (a) the name "realestate.com.au" in lower case sans serif font; (b) the silhouette of a house inside a button on a red background; (c) the colour red as the background colour on the website; and (d) the use of a tagline (which has varied from time to time) such as "Australia's No 1 property site". 35 The main brand elements (excluding the red background) are depicted in the following logo ("the realestate.com.au logo"): 36 The phrase "realestate.com.au" in the logo as shown above, (but excluding the tag line) comprise composite trade marks for which the first applicant has been the registered owner in respect of services in classes 36 and 42 since October 1999 and in respect of classes 9, 35 and 38 since September 2005. Since February 2002, the first applicant has also been the registered owner of a similar trade mark in respect of services in classes 36 and 42. In that trade mark ".com.au" is displayed in a font size much smaller than the font used to display "realestate" (collectively "the realestate.com.au trade marks"). Relevantly, class 36 services include: "Real estate affairs, including the provision of access to real estate information and analysis over a global computer network…" Class 35 includes "advertising of real estate in electronic…format". 37 In the case of REA's commercial property portal, there are also four main brand elements utilised: (a) the name "realcommercial.com.au" in lower case sans serif font; (b) a silhouette of commercial buildings inside a button on a blue background; (c) the colour blue as the background colour on the website; and (d) a tagline which often accompanies the other elements such as "Australia's No 1 commercial property site". 38 The main brand elements (without the blue background) are depicted in the following logo: 39 Since November of 2002, the first applicant has been the registered owner of a composite trade mark comprising the logo shown above (but excluding the tag line) ("the realcommercial.com.au trade mark") in respect of classes 9, 35, 36 and 42. 40 Each of the brand elements for the realestate.com.au and realcommercial.com.au portals have been extensively used by REA on the websites of those businesses where the logos appear as the title to each webpage. 41 In order to promote awareness of its brands and drive traffic to its websites, REA has promoted the realestate.com.au and realcommercial.com.au portals (including the brand elements described) in a number of different ways. Since 2005, advertising campaigns have usually been conducted in spring and autumn of each year involving online "banner" advertising as well as television, radio, cinema, print and outdoor advertising. A recent example of such an advertising campaign involved the expenditure of some $900,000. That campaign included expenditure on outdoor media spread between billboards, mobile scooters and bus shelter advertising. It also involved print media advertising in 3 metropolitan areas and some 66 radio advertisements across 4 cities. 42 Between 2005 and 2011, REA purchased more than $9m of advertising to promote the realestate.com.au portal. Between 2006 and 2011 (not including 2010 for which figures were not provided), REA purchased about $510,000 of media advertising to promote the realcommercial.com.au portal. 43 Whilst REA's advertising spend is significant, as REA's General Manager of Brand Communications and Insights Ms Joanne Whyte said, REA's advertising budget for traditional advertising is not significant compared with the media spend which might be required in order to create consumer awareness about a more traditional consumer product. In an offline environment, a great deal of advertising effort is required to build a brand in the marketplace so that consumers are familiar with it and are persuaded to consider a purchase. In the case of a property portal, the consumer has for the most part already decided to look for property online and REA's predominant objective is to bring that consumer to its website. Accordingly, REA regards the most important aspect of its advertising and promotional activities to be its search facilitation efforts. Those activities primarily comprise search engine optimisation used to enhance the exposure and ranking of REA's websites in organic search results and search engine marketing to place sponsored links to the realestate.com.au and realcommercial.com.au portals on search results pages. 44 REA has a dedicated team responsible for search engine optimisation and search engine marketing activities. That team is charged with the primary role of increasing the amount of traffic REA's websites receive from search engine searches. REA's search engine optimisation specialists design the structure of REA's web pages so that they are able to be indexed by search engines quickly and so that REA's websites are ranked highly in organic search results for the most common real estate searches. That involves ensuring that web pages that make up REA's websites contain appropriate information in their title, visible text and in the metadata that lies behind each web page, as well as in the links that connect the web page to other web pages within the group of pages that make up the websites. 45 In part because of REA's search engine optimisation activities, as at December 2011, the realestate.com.au website had 1,428,071 of its web pages indexed in Google and the realcommercial.com.au website had 580,974 pages indexed in Google. At that time, some 4,500 websites on the world wide web contained links to the realestate.com.au website. 46 REA has engaged in search engine marketing in order to promote the realestate.com.au business since at least 2005. Consultants are utilised to assist REA with search engine marketing. REA's ultimate aim is to achieve a high click through rate with a low average cost per click for terms that REA bids on. As at December 2011, REA was bidding (through AdWords) on a total of over 3.6 million keywords across 22 segmented campaigns for the realestate.com.au and realcommercial.com.au portals. 47 Between 2009 and 2011, REA's total expenditure on search engine marketing in relation to its realestate.com.au portal, including amounts paid to search engines and to consultants, exceeded $1m per annum. During that period REA's expenditure on search engine marketing for the realcommercial.com.au portal ranged between $200,000 and $300,000 per annum. 48 In the 2010-2011 financial year, the "realestate.com.au" website received more than 10 million 'clicks' from sponsored links or paid search results. Compared to the 2008-2009 financial year, REA doubled its expenditure on pay per click advertising and nearly tripled the total number of 'clicks' to the realestate.com.au website. A similar increase in both the cost of advertising and total number of 'clicks' was experienced for the realcommercial.com.au portal, which in the 2010-2011 year generated nearly 850,000 'clicks' through pay per click sponsored links. 49 REA's promotional activities include electronic marketing for consumers who have opted to receive 'email alerts'. As at January 2011 in excess of 1.7 million users were registered for email alerts via the realestate.com.au portal and almost 100,000 users were registered for email alerts via the realcommercial.com.au portal. Those users were sent in excess of 18 million emails in that month. Additionally, REA markets directly to the real estate industry, including by sending promotional material to real estate agents to provide to clients and by sponsoring industry events. 50 To promote and drive traffic to its websites, REA has also entered into various syndication agreements with owners of other popular websites. For instance, REA has a syndication agreement with eBay. Users of that website who navigate to eBay's real estate or property section are taken to a co-branded webpage which displays both eBay's name and logo alongside the "realestate.com.au" name and logo. Typically a syndication partner is paid for each unique browser that is directed to the realestate.com.au site from the syndication partner's site. Since 2006, REA has spent over $4.7m on syndication arrangements.

Real Estate 1's property portal businesses - their branding, trade marks and promotion 51 Geoffrey Luff is a director of each of the corporate respondents. He was the person primarily responsible for the establishment of the Real Estate 1 business and has been the principal decision-maker and controller of Real Estate 1's affairs. He has worked in and owned various real estate agencies in and around the Mornington Peninsula in Victoria. 52 In 2005 he decided to set up a franchise business for real estate agencies which was to be called "Real Estate 1". Part of the plan was to set up an online property portal on which real estate agents could list properties for sale. The business name "Real Estate 1" was registered in July 2006. An attempt was made at that time to obtain the domain name "realestate1.com.au" but it was unavailable. As a result Mr Luff caused to have the domain name "realestate1.net.au" registered to a company that he controlled. Mr Luff was aware of REA's realestate.com.au property portal business and understood that the property portal he intended to establish would be in direct competition with REA's portal. 53 Little or no activity to progress the establishment of the planned business occurred in 2007, other than that Mr Luff and his wife Julie (the fourth respondent) became the registered owners of a trade mark, being a composite mark in classes 36 and 42, comprising the word "realEstate" in varied font, arranged with a stylised number 1 forming part of a stylised house image, over a curved line (as shown at [55] below) ("the realestate1 trade mark"). 54 In February 2008, Mr Luff arranged for a number of domain names to be registered including "re1commercial.com.au", "realestate1commercial.com.au", "realcommercial1.com.au" and "realcommercial1.net.au". Around that time, discussions were held and work commenced on the creation of a website for Real Estate 1's property portal. Also at about that time, Mr Luff decided that he would not pursue the franchise aspect of the intended Real Estate 1 business and instead planned to focus solely on providing online advertising services through a property portal. In March 2008 the second respondent was incorporated with the intent that it would run the Real Estate 1 business. 55 On 30 October 2008 Real Estate 1 launched a portal for listings of residential property for sale and rent with the domain name "realestate1.net.au". At the same time, a further website was launched with the domain name "commercial1.net.au". This portal was intended for listings of commercial real estate for sale or lease. On each page of the residential property website the realestate1 trade mark appeared as a logo in the title of the page. The logo included the words "realEstate1" accompanied by a silhouette of a house in an orange colour over an orange coloured curved line, as follows: 56 A similar logo (for which Real Estate 1 has not held a registered trade mark) in which the term "commercial1" appeared, was utilised in the title of Real Estate 1's commercial property portal, as follows: 57 In November 2008, Mr Luff acquired the domain name "realestate1.com.au" and the second respondent became the registrant of that domain name. 58 In March 2009, Real Estate 1's residential property portal was (to use Mr Luff's words) "re-branded as realestate1.com.au". The logo at the top of the web pages on the website was altered so that the words ".com.au" appeared following the silhouette of the house as shown below: 59 This logo was used by Real Estate 1 for its residential property portal between March 2009 until on or about the commencement of this proceeding on 21 October 2010. During that time, the logo appeared prominently on the title of each webpage on the website. Also during that time, the text on the website variously referred to the business operating the website as "realEstate1" or "Real Estate One" or "realestate1.com.au". 60 In the period March 2009 to about October 2010, the logo set out at [58], was also used in advertising and promotional material for Real Estate 1's residential property portal. The evidence suggests that about $50,000 was expended by Real Estate 1 on traditional advertising (mainly radio advertising) during the period. Additionally, Real Estate 1 produced brochures, key rings and bumper stickers for distribution to agents and their clients. Forms distributed to agents also contained the "realestate1.com.au" logo as did the signage on the vehicle used by Real Estate 1's sales representative. Real Estate 1 regarded the use of its domain name "realestate1.com.au" as vital for its advertising, given that it was trading online. 61 On 23 November 2009, the first respondent was incorporated with the intent that it would take over the Real Estate 1 business. A number of investor presentations were made at or about that time to which I will later refer. Offer documents and a prospectus issued to prospective investors in the first respondent also utilised the logo set out at [58]. 62 In response to this litigation and from about late October 2010, the ".com.au" was removed from the logo appearing in the title of web pages on Real Estate 1's residential property portal. The logo used reverted to a mark consistent with the realestate1 trade mark. 63 However, Real Estate 1 continued to use the domain name "realestate1.com.au" and operate its residential property portal from a URL containing that domain name. Beyond its use as part of an internet address, there is evidence of some minor continued use of the name "realestate.com.au". The signage on Real Estate 1's motor vehicle was not altered; the logo containing the domain name continued to appear in some minor references to Real Estate 1's residential property portal made on Real Estate 1's commercial property portal; and on one page on the "realestate1.com.au" website a reference is made to the domain name which suggests that the website is operated by a business of that name. A similar reference equating the name of the operator of the business to its domain name, is made on a web page for Real Estate 1's residential property portal that is accessible to users of mobile phones. 64 Between August 2009 and about late October 2010 (when the proceeding was commenced), Real Estate 1's commercial property portal was conducted at a URL address which included the domain name "realcommercial1.com.au" and was otherwise carried out under the name "realcommercial1.com.au" by reference to a logo which incorporated that name accompanied by a silhouette of a commercial building on a blue background over a curved line, as follows: 65 The logo was used in the title of each of the web pages on Real Estate 1's commercial property portal. In response to this litigation and from about late October 2010, the domain name for Real Estate 1's commercial property portal was changed to "realestate1commercial.com.au" and the branding on the website was altered to "realEstate1Commercial" with a logo as follows: 66 Within about three weeks of the launching of Real Estate 1's property portals in late 2008, some 150 real estate agents had registered with Real Estate 1. Those numbers increased over time. By December 2010, there were 1,036 agents registered. As at February 2012, there were 1875 agents registered. Agents registered with Real Estate 1 may either take up a free or a paid subscription. An agent with a free subscription can upload an unlimited number of properties to the website. A paid subscription (at a cost of $100 per month) provides access to additional features and tools. The extent to which real estate agents pay for subscriptions was not identified in the evidence. 67 In the period 1 April 2009 to 31 January 2012, just over 900,000 visits were made to the realestate1.com.au website. Some 78% of that traffic came from search engines. Real Estate 1 has engaged in search engine marketing since at least April 2009 and search engine optimisation since July 2009. 68 Despite those efforts, and whilst I accept that it is a competitor of REA, Real Estate 1 must be regarded as a relatively insignificant participant in the property portal market. In the month of October 2011, when realestate.com.au ranked as the number one property portal in Australia with over 16 million visits representing some 34% of total visits to property portals, the realestate1.com.au website ranked at 121 with about 25,500 visits representing 0.0544% of total visits to property portals. Financial reports for Real Estate 1 show that in the financial year ended 30 June 2011 it had a gross profit from trading of about $325,000 and a net operating loss of about $182,000.

REA's Pleaded Claims 69 REA pleaded that the domain name "realestate1.com.au" is substantially identical or deceptively similar to the domain name "realestate.com.au" and that the domain name "realcommercial1.com.au" is substantially identical or deceptively similar to the domain name "realcommercial.com.au". It also pleaded that the trade indicia by reference to which Real Estate 1's residential and commercial property portal businesses have been conducted since March 2009 (mostly detailed at [58]-[65] above), are substantially identical or deceptively similar to the trade indicia by which the realestate.com.au and realcommercial.com.au businesses have been conducted (mostly detailed at [34]-[39] above). 70 By reference to the use (on its websites, in advertising and in promotional materials) by Real Estate 1 of names and trade indicia that are claimed to be substantially identical or deceptively similar to those employed by REA, REA pleaded that Real Estate 1 has and continues to make false representations to the public which may be summarised as follows: (a) Real Estate 1's residential property portal business is a business of, or service of, or associated with, authorised or approved by, REA; (b) Real Estate 1's commercial property portal business is a business of, or service of, or associated with, authorised or approved by, REA; (c) the realestate1.com.au website is a website of, or associated with, authorised or approved by, REA; and (d) the realcommercial1.com.au website is a website of, or associated with, authorised or approved by, REA. 71 By reason of that conduct, REA pleaded that Real Estate 1 had: (a) engaged in conduct that is misleading or deceptive or likely to mislead or deceive, in contravention of section 52 of the TPA; (b) represented, in connection with the supply of its services, that those services have the sponsorship and approval of REA when they do not have such sponsorship or approval, in contravention of section 53(c) of the TPA; and (c) represented that it has the sponsorship or approval of, or an affiliation with REA that it does not have, in contravention of section 53(d) of the TPA. 72 Additionally, by reason of that same conduct, REA pleaded that Real Estate 1 had passed off: (a) its residential property portal business as a business of, or service of, or associated with, authorised or approved by, REA; (b) its commercial property portal business as a business of, or service of, or associated with, authorised or approved by, REA; (c) the realestate1.com.au website as a website of, or associated with, authorised or approved by, REA; and (d) the realcommercial1.com.au website as a website of, or associated with, authorised or approved by, REA. 73 REA complained that by its conduct Real Estate 1 had injured, and if not restrained will continue to injure, the goodwill and reputation of REA in relation to: (i) the domain name "realestate.com.au"; (ii) the domain name "realcommercial.com.au"; (iii) the trade indicia utilised by the realestate.com.au business; and (iv) the trade indicia utilised by the realcommercial.com.au business. 74 As I will later explain, REA's case as expounded at trial was much narrower than that pleaded. At this point I should say that the claimed contraventions of s 53 of the TPA were not pressed as other than supplementary to the s 52 claim and on the basis that their fate would be no different to that of the s 52 claims. The parties approached the s 52 claims and the claims of passing off in the same way. REA did not contend that there was a basis for it succeeding on its passing off claims if it failed on its s 52 case. A range of issues common to both the s 52 and passing off claims arise for determination and I have dealt with them in tandem but have expressed my consideration of them by reference to the s 52 claims.

The relevant statutory provisions 75 At the time of the commencement of this proceeding, the applicable legislative provisions were contained in the TPA. In late 2010, the title of the TPA and a number of its provisions were amended by the Trade Practices Amendment (Australian Consumer Law) Act (No 2) 2010 (Cth) and its short title became the Competition and Consumer Act 2010 (Cth). The transitional arrangements provide that the TPA provisions continue to apply to conduct occurring prior to 1 January 2011. 76 As outlined above, REA alleges that the respondents have engaged in conduct that contravenes ss 52, 53(c) and 53(d) of the TPA. Those provisions are in the following terms: 52 Misleading or deceptive conduct (1) A corporation shall not, in trade or commerce, engage in conduct that is misleading or deceptive or is likely to mislead or deceive. (2) Nothing in the succeeding provisions of this Division shall be taken as limiting by implication the generality of subsection (1). 53 False or misleading representations A corporation shall not, in trade or commerce, in connexion with the supply or possible supply of goods or services or in connexion with the promotion by any means of the supply or use of goods or services: ….. (c) represent that goods or services have sponsorship, approval, performance characteristics, accessories, uses or benefits they do not have; (d) represent that the corporation has a sponsorship, approval or affiliation it does not have;

Some general principles 77 The principles underpinning s 52 are for the most part well established and were not in dispute. Some general principles are dealt with here, whilst principles requiring particular consideration are dealt with later in these reasons. 78 Section 52 establishes a standard of conduct that traders must adhere to in their dealings with consumers. A rival trader, consumers and the regulator are each entitled to bring an action in respect of a contravention of this standard. 79 Conduct which causes confusion or wonderment will not necessarily be misleading or deceptive: Google Inc v Australian Competition and Consumer Commission (2013) 294 ALR 404 at [8] (French CJ, Crennan and Kiefel JJ). If the conduct of a corporation gives rise to confusion and uncertainty in the minds of the public about whether two products or services might have come from the same source, the corporation does not necessarily contravene s 52: Parkdale Custom Built Furniture Pty Ltd v Puxu Pty Ltd (1982) 149 CLR 191 at 200 (Gibbs CJ); Taco Company of Australia Inc v Taco Bell Pty Ltd (1982) 42 ALR 177 at 201 (Deane and Fitzgerald JJ); Bridge Stockbrokers Ltd v Bridges (1984) 4 FCR 460 at 472-473 (Lockhart J). However, if the conduct of a corporation causes more than mere confusion and causes consumers to actually conclude that two products do come from the same source, such conduct is likely to be misleading and deceptive: Bridge Stockbrokers at 473 (Lockhart J). The representee must be led into error and labour under an erroneous assumption: Campomar Sociedad Limitada v Nike International Limited (2000) 202 CLR 45 at [104] (the Court). 80 The provision does not preclude a rival trader from entering the market and copying the product of another in an attempt to take some of its market share (subject to any infringement of property rights), provided the rival does not mislead or deceive consumers or pretend that its goods are the goods of another: Peter Bodum A/S v DKSH Australia Pty Ltd (2011) 280 ALR 639 at [202] (Greenwood J, with whom Tracey J agreed); citing Interlego AG v Croner Trading Pty Ltd (1992) 39 FCR 348 at 391-392 (Gummow J, with whom Black CJ and Lockhart J agreed). 81 The High Court in Google Inc explained that the words "likely to mislead or deceive" make clear that it is not necessary to demonstrate actual deception to establish a contravention of s 52: Google Inc at [6] (French CJ, Crennan and Kiefel JJ). Evidence of actual deception may be persuasive but is not essential, the test is objective and the court must determine for itself whether conduct is misleading or deceptive or likely to mislead or deceive: Taco Bell at 202 (Deane and Fitzgerald JJ). 82 In cases involving representations made to an indeterminate number of people, the court must first determine the class of people to whom the representation was made and the likely attributes of the 'ordinary' or 'reasonable' members of that class. The High Court in Campomar explained (at [102]-[103]): It is in these cases of representations to the public, of which the first appeal is one, that there enter the "ordinary" or "reasonable" members of the class of prospective purchasers. Although a class of consumers may be expected to include a wide range of persons, in isolating the "ordinary" or "reasonable" members of that class, there is an objective attribution of certain characteristics. Thus, in Puxu, Gibbs CJ determined that the legislation did not impose burdens which operated for the benefit of persons "who fail[ed] to take reasonable care of their own interests". In the same case, Mason J concluded that, whilst it was unlikely that an ordinary purchaser would notice the very slight differences in the appearance of the two items of furniture in question, nevertheless such a prospective purchaser reasonably could be expected to attempt to ascertain the brand name of the particular type of furniture on offer. Where the persons in question are not identified individuals to whom a particular misrepresentation has been made or from whom a relevant fact, circumstance or proposal was withheld, but are members of a class to which the conduct in question was directed in a general sense, it is necessary to isolate by some criterion a representative member of that class. The inquiry thus is to be made with respect to this hypothetical individual why the misconception complained has arisen or is likely to arise if no injunctive relief be granted. (Citations omitted) See further, Google Inc at [7] (French CJ, Crennan and Kiefel JJ). 83 The initial question that must be determined is whether the misconceptions or deceptions alleged to arise or to be likely to arise, can properly be attributed to the 'ordinary' or 'reasonable' members of the class: Campomar at [105] (the Court). 84 A different approach is to be taken where the target of the impugned conduct is an identified individual: Butcher v Lachlan Elder Realty Pty Ltd (2004) 218 CLR 592 at [36] (Gleeson CJ, Hayne and Heydon JJ); Eatock v Bolt (2011) 197 FCR 261 at [244]-[246] (Bromberg J). 85 A number of judges of this Court have held that in order for conduct to breach s 52, it must be established that a "not insignificant" proportion of people to whom representations were made, would be misled or deceived by the impugned representation. In .au Domain Administration Ltd v Domain Names Australia Pty Ltd (2004) 207 ALR 521 at [21]-[26], Finkelstein J listed the relevant authorities and traced that requirement to the law of passing off and its consequent application to s 52. The judge was critical of that approach including because he thought it inconsistent with the test adopted by the High Court in Campomar: see discussion in Bodum at [206]-[210] (Greenwood J, with whom Tracey J agreed). In the absence of express High Court authority I am bound to follow the approach expressed by Greenwood J (with whom Tracey J agreed) in Bodum. Applying this approach, it appears that once certain objective characteristics have been attributed to 'ordinary' or 'reasonable' members of the class, there remains scope for some diversity of response to the impugned conduct. It is then necessary to ask whether a "not insignificant" proportion of 'ordinary' or 'reasonable' members of the class were misled or likely to be misled. 86 Section 52 is not directed to the intention of a party and it is possible for a contravention to occur even where a party has acted reasonably and honestly: Google Inc at [9] (French CJ, Crennan and Kiefel JJ); Hornsby v Building Information Centre Pty Ltd v Sydney Building Information Centre Ltd (1978) 140 CLR 216 at 228 (Stephen J).

Is the impugned conduct that of Real Estate 1? 87 An assessment of whether conduct is misleading or deceptive or likely to be so, should logically commence with the identification of the impugned conduct. Where the impugned conduct is a representation, it is necessary to determine whether the representation was made by the alleged contravener. The conduct here in question is clear. It includes, as I have indicated already, the use of either "realestate1.com.au" or "realcommercial1.com.au" displayed in sponsored links and organic search results which relate to the property portals of Real Estate 1. 88 Although the matter was not the subject of direct evidence and was only dealt with in general terms, there was no issue that the content of sponsored links paid for by Real Estate 1 was content created by Real Estate 1 and that the representations there made was conduct of Real Estate 1. Nor was it suggested by Real Estate 1 that the content of organic search results for its property portals was not authored by it or that representations there contained were not the conduct of Real Estate 1. Although the heading, URL and page description which appear in an organic search result are technically determined by the search engine, Real Estate 1 includes on its web pages metatags for both the heading and page description designed to facilitate the adoption and display of that content in organic search results.

The relevant class 89 To determine whether Real Estate 1's conduct was misleading or deceptive, it is necessary to make findings about how consumers would have perceived that conduct including, whose conduct consumers would have perceived the conduct to be and what, if any, representations would have been perceived to flow from the conduct. 90 In order to carry out that task, it is necessary to identify the class of consumers to whom the impugned representations were made. Further, because the Court must consider whether the 'ordinary' or 'reasonable' members of that class would be misled or deceived, it is necessary to identify the class by reference to the characteristics of the 'ordinary' or 'reasonable' persons in the class. 91 Whilst the identity of the relevant class is fairly straightforward, the determination of the characteristics which ought to be ascribed to the 'ordinary' or 'reasonable' persons in the class is more difficult and more crucial to the result that I have reached. I will identify shortly some general characteristics which I consider should be ascribed and deal later with further specific characteristics that are of particular relevance. 92 In determining the characteristics that ought to be ascribed to the 'ordinary' or 'reasonable' persons in the class, I have been somewhat assisted by evidence called by REA from Brent Coker. Mr Coker is a lecturer in Marketing at the University of Melbourne and REA relied upon his evidence as the opinion of an expert. He holds a PhD in Information Systems and Electronic Commerce amongst other qualifications. Mr Coker has been involved in extensive research and I am satisfied has some expertise in online consumer behaviour and internet marketing. The evidence which Mr Coker gave about online consumer behaviour was only partly based on formal research. Much of Mr Coker's evidence and all of the evidence given by Phillip Leahy (an online marketing and sales professional with 15 years experience in online marketing, who was called by Real Estate 1) was based on the witness's familiarity with the nature and content of a search results page and what the witness regarded as the likely plausible reactions of consumers, given the type of search engaged in and the information likely to be presented on the search results page. 93 Having become acquainted, through the evidence, with the way in which search engines operate and with the nature and likely content of information displayed on search engine results pages, the Court was as well placed as were Mr Coker and Mr Leahy to make judgments about the likely plausible behaviour of consumers. I have some doubt as to the extent to which much of the evidence given by Mr Coker and all of Mr Leahy's evidence, can properly be characterised as expert evidence. That is not to say that the Court was not assisted by the evidence given, in the same way that the Court may be assisted by submissions. I was assisted by some of Mr Coker's evidence although I have not accepted many of his conclusions. In the main that is because Mr Coker's views were given without taking into account many surrounding facts of importance to which I will later refer. To some extent, Mr Leahy's conclusions were similar to those made by Mr Coker. To some extent their conclusions differed. Objection was taken to the expertise of Mr Leahy and the admissibility of his evidence. As I was not assisted and I have not relied upon conclusions made by Mr Leahy, it is not necessary that I determine the objection. 94 The characteristics to be attributed to the 'ordinary' or 'reasonable' persons in the class which the High Court considered in Google Inc, were findings made at first instance by Nicholas J in Australian Competition and Consumer Commission v Trading Post Australia Pty Ltd (2011) 197 FCR 498. Nicholas J had to determine whether representations made in sponsored links were misleading or deceptive. Given the similarity of the consumer transactions dealt with in that case to those in this case, I have also been assisted by the characterisation of the relevant class and their attributes as found by Nicholas J in Trading Post. 95 The identification of the relevant class was not particularly controversial. The class consists of people who have access to a computer connected to the internet and who use search engines to view real estate listings on property portals. The class, therefore is actual or prospective users of REA's property portals: Bodum at [205] (Greenwood J, with whom Tracey J agreed). 'Ordinary' or 'reasonable' persons within that class will have at least a basic knowledge and understanding of computers, the internet and of how to use search engines. 96 The level of familiarity of such persons with the functionality of search engines and the nature and content of a search results page means that such persons are likely to understand that: (i) search results include advertisements (which are identified as such) and other results; (ii) the advertisements or 'sponsored links' will, like all advertisements, be paid for by the advertiser, contain content provided by the advertiser and are likely to have a prominence of place on the search page referrable to the payment made by the advertiser; (iii) the other results (the organic search results) are generated and ordered by the search engine predominantly by reference to the relevance of the websites displayed to the keywords searched for by the searcher; and (iv) the content of the organic search results has likely been taken by the search engine from the web page or website whose URL is displayed in the result and as such, the words displayed are likely to be those of the web page and not words authored by the search engine. 97 As a result of the consumers' understanding set out at (ii) and (iii) above, the 'ordinary' or 'reasonable' persons in the class will understand that an organic search result is likely to be more relevant to the information searched for than a sponsored link, and that the most relevant organic results will appear in order of relevance commencing with the first result. In that latter respect, I accept Mr Coker's evidence which included reference to research that demonstrated that a consumer will click onto the first organic search result displayed on a search results page on some 42% of occasions and that the first five positions have approximately a 76% chance of being chosen, and that 90% of consumers do not visit page 2 or beyond. 98 By reason of (i) and (ii) above, the 'ordinary' or 'reasonable' member of the class will perceive that representations made in or by a sponsored link are representations of the advertiser. At least in respect of on-line businesses which trade through the internet, the 'ordinary' or 'reasonable' person will perceive a representation made in or by an organic search result relating to such a business, as a representation made by the business concerned. 99 I also ascribe to the ordinary and reasonable members of the class, a basic understanding of general internet conventions. This includes an understanding that ".com" is commonly used by commercial entities and ".au" by Australian based entities, as well as knowledge that every web page has a unique URL and that any variation between one URL and another is significant and indicative of a different web page or website.

Are REA's names distinctive of its businesses? 100 Before going on to consider whether a not insignificant proportion of the 'ordinary' or 'reasonable' members of the relevant class were likely to have been misled or deceived, it is necessary to say something about the distinctiveness of the REA domain names which are used as the trading names for its residential and commercial property portals and to consider whether these names have acquired a secondary meaning. 101 As outlined earlier, REA alleges that Real Estate 1 adopted names and trade indicia substantially identical or deceptively similar to those of REA's property portal businesses, such that Real Estate 1 represented that its property portals were affiliated or connected with REA. To make good this allegation, REA must establish that its names have acquired a secondary meaning and have become distinctive of its businesses. After all, consumers who are not familiar with "realestate.com.au" as the name for REA's residential property business could not possibly be misled or deceived into believing that a reference to "realestate1.com.au" is a reference to the business of realestate.com.au or an associated business. Therefore, the extent to which consumers associate the names "realestate.com.au" and "realcommercial.com.au" with REA's online property portal businesses is an important issue in this case. 102 No evidence was called by REA directly from consumers in relation to consumer familiarity with the name, brand or reputation of REA's portals. Instead, REA relied on the promotion of its brands and the very substantial use of its websites by consumers, to establish a substantial reputation in its property portals and their association in the minds of consumers with REA's property portal businesses. 103 That REA has established substantial brand recognition by consumers for its residential and commercial real estate portals is not in issue. Real Estate 1 conceded that by the use of the combination of the brand elements used to promote REA's residential and commercial real estate portals, REA has developed a reputation in the Australian market. It was not suggested that the combination of those brand elements were not, in each case, sufficient to be distinctive of REA's portal businesses. 104 Real Estate 1 relevantly contended that what REA had failed to establish was that the term "realestate.com.au", when appearing on a search results page, had become distinctive of REA's business. Given the specific nature of the impugned conduct expounded at trial (see further at [136]-[137] below), REA needed to establish that when "realestate.com.au" appeared in a search result, that term was distinctive of REA's residential portal business. 105 It is difficult to establish that a descriptive name, as opposed to a concocted or invented name, is a name which has become distinctive of a trader's business: Cellular Clothing Co Ltd v Maxton & Murray [1899] AC 326 at 336 (Earl of Halsbury LC). Whether a name is distinctive of a particular business is a question of fact and degree: Architects (Australia) Pty Ltd v Witty Consultants Pty Ltd [2002] QSC 139 at [19] (Chesterman J). A concocted name is far more likely to be distinctive of a business than a name which is based upon a description of the nature of the business that is being conducted: Chase Manhattan Overseas Corporation v Chase Corporation Ltd (1985) 9 FCR 129 at 147 (Wilcox J). Whilst a name may at the same time be both descriptive and distinctive, the fact that a name prima facie retains its descriptive signification increases the difficulty of proving that it is distinctive of the goods of a particular business: Burberrys v JC Cording & Co (1909) 26 RPC 693 at 704 (Parker J); British Vacuum Cleaner Co Ltd v New Vacuum Cleaner Co Ltd [1907] 2 Ch 312 at 322-323 (Parker J); and South Australian Telecasters Limited v Southern Television Corporation Limited [1970] SASR 207 at 220 (Walters J). 106 In Kettle Chip Co Pty Ltd v Apand Pty Ltd (1993) 46 FCR 152 at 165-6, Burchett J stated: "Of course, the mere fact that a word is descriptive does not warrant a trader in using it if it has also acquired a secondary meaning as to distinguish the products of a rival. As Lord Radcliffe said in de Cordova v Vick Chemical Co (supra) at 106: 'To say that it is descriptive would not be enough, for…there is no absolute incompatibility between what is descriptive and what is distinctive. A descriptive word, such as 'Sheen', can be recognised in law as distinctive if the evidence clearly shows that it is distinctive in fact". 107 In Cantarella Bros Pty Ltd v Modena Trading Pty Ltd [2013] FCA 8 at [108] and by reference to the judgment of Lockhart J in Johnson & Johnson Australia Pty Ltd v Sterling Pharmaceuticals (1991) 30 FCR 326, Emmett J set out what is meant by the requirement that a term be "distinctive". Whilst the analysis was given in the context of a trade mark case, it bears repeating here: There must be a sufficient degree of distinctiveness to counterbalance the descriptive character of a word. A word that is prima facie descriptive may become distinctive in connection with particular goods but yet retain its descriptive meaning. However, the word must, to become distinctive, have a new and secondary meaning different from the descriptive one, and thus cease to be purely descriptive. Distinctive means that the mark distinguishes the registered proprietor's goods from others of the same type in that market. However, distinctive does not mean that the goods must specifically identify the registered proprietor as the source of the goods. What is important is that a significant number of consumers in the relevant market identify the registered proprietor's goods as coming from one trade source (Johnson & Johnson at 335-336). 108 Examples of names which have been held to have achieved the requisite secondary meaning have included "Minties" (for mints): Angelides v James Stedman Hendersons Sweets Ltd (1927) 40 CLR 43 at 72; "Budget Rent a Car" (for car rental services): BM Auto Sales Pty Ltd v Budget Rent a Car System Pty Ltd (1976) 12 ALR 363 at 369; "Opals Australia" (for opal sales): Opals Australia Pty Ltd v Opal Australiana Pty Ltd (1993) ATPR 41-264; "Microwave Cuisine" (for microwave cooking classes): LSK Microwave Advance Technology Pty Ltd v Rylead Pty Ltd (1989) 16 IPR 107 at 114; and "Architects Australia" (for architectural services): Architects Australia at [15]-[24]. In each case, the evidence before the court showed that the names, though descriptive, had come to be associated by consumers with the goods or services of a particular trader. Examples of names which have been held not to achieve the requisite secondary meaning have included "Flexibond" (for an investment product): Lumley Life Ltd v IOOF Friendly Society (1989) 16 IPR 316; "Cellular" (for a type of cloth): Cellular Clothing; "Office Cleaning Services" (for a cleaning business): Office Cleaning Services Ltd v Westminster Window and General Cleaners (1946) 1 All ER 320; and "British Vacuum Cleaner Company" (for a company that held a patent for particular vacuum cleaner technology): Vacuum Cleaner Case. Whilst each of these cases is illustrative of the relevant principles, each turns on its particular facts. The facts of this case support a finding that REA's domain names have become distinctive, in the sense that they distinguish REA's property portals from other portals in the same market. 109 The very substantial efforts engaged in by REA to advertise and promote its portal businesses and the very substantial success of those businesses make it abundantly clear that the businesses are both popular and well-known to a significant number of Australian consumers of property portal services. Whilst the logos and other brand elements that are used to promote the businesses contribute significantly to consumer recognition of those businesses, I accept REA's contention, including the evidence of Mr Coker, that (in each case) the name is the dominant brand element. Whilst I find in the following part of this judgment that as a term, "realestate.com.au" is only faintly recognisable as an intended name for a brand or business, nonetheless I consider that the nature and extent of the advertising of the term in connection with REA's residential portal, makes it likely that when "realestate.com.au" is displayed as a trading or domain name for that portal, it is recognised as such by a significant number of Australian consumers who use property portals. Reputation may be inferred from a high volume of consumer use, together with substantial advertising expenditure and promotion: McCormick & Co Inc v McCormick [2000] FCA 1335 at [86]-[88] (Kenny J); Playcorp Group of Companies Pty Ltd v Peter Bodum A/S (2010) 84 IPR 542 at [84] (Middleton J). 110 The manner in which consumers reach REA's websites also strongly favours a finding that REA's domain names are distinctive of its property portal businesses because those names are utilised by consumers when trying to locate those businesses. Analytical tools capable of tracking and recording the source of traffic coming to a website have been utilised by REA to show the source of internet traffic to REA's websites. In the first six months of 2011, some 53% of visits to the realestate.com.au website came from direct traffic. Direct traffic refers to those visits made to a website when an internet user has typed the web address for the website into the user's web browser, or has clicked a saved link to that web address (for example, a link saved as a 'favourite' link). Access through a saved link does not evidence name recognition but other access via a web browser does. Whilst no breakdown was available which identified the proportion of direct traffic that arrived directly from a web browser, I would infer that a very significant number of consumers have gained access in that way. Given the very substantial numbers overall, there must be many hundreds of thousands of consumers in that category. 111 More specific evidence was produced that suggests very substantial consumer familiarity with the term "realestate.com.au". In July 2010, the most popular search term driving traffic to the realestate.com.au website from organic search results was the term "realestate.com.au". Some 6% of all traffic arriving at the realestate.com.au website arrived as a result of a searcher using that keyword phrase. Between 9 November 2011 and 9 December 2011, on about 1 million occasions, consumers came to the realestate.com.au website having clicked on an organic search result that was displayed because the consumer used "realestate.com.au" as the keyword search term. In the same period, on about 110,000 occasions, consumers arrived through a similar path having used "www.realestate.com.au" as a keyword search term. 112 That "realcommercial.com.au" is a term which consumers associated with REA's commercial property portal was not in contest. Although the term "realcommercial" is made up of two descriptive words, its nature differs from that of "realestate.com.au". A series of descriptive words may be so organised as to create a unique and distinctive name for a product, as the "Healthy Fishing-Rod" example given by Isaacs J in Angelides at 62 demonstrates. When the words "real" and "commercial" are combined, a concocted word is formed by the combination. A concoction is rarely plain and usually memorable, including because the creation of something new will be thought to be clever or inspiring or at the very least unique. It is therefore not very difficult to establish that a concocted name used in connection with a business has acquired, in the minds of consumers, an association with that business. 113 The term "realcommercial.com.au" is not strongly descriptive of a property portal in the way that "realestate.com.au" is. The addition of ".com.au" to the concocted word will be readily recognised as merely indicating the online focus of the business concerned. 114 The extent of advertising and promotion of REA's realcommercial.com.au website was sufficiently apparent from the evidence to lead me to infer that it has been significant if not substantial. The popularity of the site confirms its extensive use and strongly suggests that the business has strong name recognition amongst consumers interested in commercial real estate. There was also evidence of the use of the name "real commercial" by consumers when trying to locate the realcommercial.com.au website. In July 2011, the keyword search term "real commercial" was one of the two leading search terms driving traffic to the realcommercial.com.au website. It accounted for about 2% of the traffic arriving at realcommercial.com.au. The search term "realcommercial.com.au" also accounted for 0.7% of traffic. 115 It is clear a significant number of consumers who utilise search engines to search for property portals, recognise REA's residential property portal by the name "realestate.com.au". This is also the case in relation to REA's commercial property portal and the name "realcommercial.com.au". Accordingly, I am satisfied that secondary meaning has been established in the terms "realestate.com.au" and "realcommercial.com.au" and that a significant number of consumers associate these terms with REA's online property portal businesses.

Causation and the extent of differentiation required to avoid deception 116 When a descriptive name is used to identify a brand or business, any misconception on the part of consumers may be the result of the indistinct name adopted by the business rather than any conduct on the part of its competitor. Further, in relation to the use of descriptive names, small differences in the names used by competitors will suffice to avoid misconception by consumers. The principles established in Hornsby bear heavily on these questions and need now to be discussed before I turn more specifically to the question of whether 'ordinary' or 'reasonable' members of the relevant class were likely to have been misled. In particular, I have in mind what Stephen J said on the issue of causation and the use of descriptive words as trading names. In Hornsby, Stephen J (with whom Jacobs J agreed) delivered the leading judgment. Barwick CJ (with whom Aickin J agreed), also expressed agreement with Stephen J, insofar as Stephen J's reasoning supported the conclusions of Barwick CJ. 117 Why misconception has arisen in the minds of consumers was the springboard question posed by Stephen J in Hornsby. That focus upon causation was endorsed by the High Court in Campomar at [98] in requiring that there be established a "sufficient nexus" between the impugned conduct and any misconception or deception: see also Bing! Software Pty Ltd v Bing Technologies Pty Ltd (2009) 180 FCR 191 at 204 (Kenny J, with whom Greenwood and Logan JJ agreed); Dairy Vale Metro Co-operative Ltd v Brownes Dairy Ltd (1981) 35 ALR 494 at 501-502 (Toohey J); Snyman v Cooper (1989) 91 ALR 209 at 226 (von Doussa J). In Taco Bell at 203 Deane and Fitzgerald JJ said: Finally, it is necessary to inquire why proven misconception has arisen: Hornsby Building Information Centre v Sydney Building Information Centre (18 ALR at 647 140 CLR at 228). The fundamental importance of this principle is that it is only by this investigation that the evidence of those who are shown to have been led into error can be evaluated and it can be determined whether they are confused because of misleading or deceptive conduct on the part of the respondent. 118 By reference to that causative requirement, Stephen J identified a need, in the context of a claim of misleading or deceptive conduct based upon a similarity between the name utilised by the applicant and that of its rival, to consider the contribution made by the applicant's adoption of a name consisting of descriptive words. In a well known and often cited passage, Stephen J said at 229: There is a price to be paid for the advantages flowing from the possession of an eloquently descriptive trade name. Because it is descriptive it is equally applicable to any business of a like kind, its very descriptiveness ensures that it is not distinctive of any particular business and hence its application to other like businesses will not ordinarily mislead the public. In cases of passing off, where it is the wrongful appropriation of the reputation of another or that of his goods that is in question, a plaintiff which uses descriptive words in its trade name will find that quite small differences in a competitor's trade name will render the latter immune from action: Office Cleaning Services Ltd v Westminster Window and General Cleaners Ltd (1946) 63 RPC 39, per Lord Simonds at 42. As his Lordship said (at 43), the possibility of blunders by members of the public will always be present when names consist of descriptive words: "So long as descriptive words are used by two traders as part of their respective trade names, it is possible that some members of the public will be confused whatever the differentiating words may be." The risk of confusion must be accepted, to do otherwise is to give to one who appropriates to himself descriptive words an unfair monopoly in those words and might even deter others from pursuing the occupation which the words describe. 119 Stephen J considered that these principles applied equally to s 52 of the TPA including that "quite small differences" will suffice where descriptive words are used in trade names. His Honour reasoned that "the creation of any monopoly in descriptive names would be to mock the manifest intent of the legislation" (at 230). 120 In Hornsby, an entity which traded as the "Sydney Building Information Centre" claimed that the name "Hornsby Building Information Centre" used by a rival was misleading and deceptive. Stephen J determined that any deception that arose in the circumstances of that case stemmed, not so much from the rival's use of a similar name, but from the applicant's adoption of descriptive words in its own name. His Honour determined that the confusion created in the minds of the public was not anything to which s 52 was directed (at 230). 121 REA's suggestion that the reasoning in Hornsby had now been somewhat overtaken was not supported by authority. The reliance upon it in Campomar denies that conclusion. I regard Hornsby as both binding and persuasive. It provides "guidance as to the general approach to be adopted in s 52 cases involving the use of similar names": Taco Bell at 201 (Deane and Fitzgerald JJ). 122 In Connect.com.au Pty Ltd v GoConnect Australia Pty Ltd (2000) 178 ALR 348 at [57]-[61], Emmett J applied Hornsby in relation to the word "connect". The judge noted that where an ordinary word acquires a secondary meaning "very little" differentiation is required in order to distinguish its use for the purposes of s 52. 123 The words "real estate" are undeniably descriptive words. A business involved in the sale or rent of real estate under the name "Real Estate" alone has no prospect of establishing that the words "real estate" are distinctive of its business. REA accepts that "real estate" is purely descriptive. REA contended however that the combination of "real estate" with ".com.au" created a term which is distinctive of REA's residential property portal. 124 There are a number of features about the term "realestate.com.au" and the context in which it is used, which tend to support the conclusion that only a small difference in a competitor's trade name "will render the latter immune from action" (Hornsby at 229). 125 The first is that each of the elements of the term "realestate.com.au" remain descriptive despite the combination of them into a single term. The phrase "real estate" is a very well known and frequently used phrase. Whilst there are other ways in which to describe real property, there are not many and "real estate" is the description most commonly used. A "dot-com" is a company which conducts business on the internet and, in an internet address, ".com" indicates a commercial internet site: Shorter Oxford English Dictionary (6th ed, Oxford University Press, 2007) p 744. The term is obviously derived from the word "commercial" and it is the most common top-level domain used by commercial businesses. The top-level domain ".au" is a well known abbreviation for Australia and, when used as part of an internet address, it is descriptive of the Australian location of the website concerned. 126 Secondly, the phrase "real estate" will not be regarded as a name in the sense of being a sign or badge of origin or identity. Real estate is a common English phrase which suggests no necessary connection with the existence of a business. It is a descriptive word which describes articles of trade and not the origin of those articles: Cellular Clothing at 337 (Lord Watson). It is only when ".com.au" is attached to "realestate" and the term "realestate.com.au" is viewed as a single expression, that the term begins to suggest a connection with a source business or brand. However, the very features (.com.au) which are essential for giving the term recognition as a name, are features which the cases indicate will commonly be ignored or be given little substantive attention by consumers. 127 In determining whether one name suggests the business of another, the subsidiary elements of a name are generally not regarded as significant. Thus in Hornsby, the locality word ("Sydney") which prefixed the three descriptive words was given no particular weight. The comparison made was focused upon the dominant descriptive words alone. Similarly, in British Diabetic Association v Diabetic Society Limited, the prefix "British" was "put on one side" in the comparison made between "British Diabetic Association" and "Diabetic Association": [1995] 4 All ER 812 at 831 (Robert Walker J). 128 The lack of attention which may be paid to a suffix such as ".com.au" is illustrated in Architects Australia. In that case a comparison was made between the names "Architects Australia" and "architectsaustralia.com.au". To the suggestion that ".com.au" differentiated the two names, Chesterman J said at [21]: The words in contest in this case are not merely similar, they are precisely the same. The extended suffix ".com.au" does not detract from this conclusion. It does no more than indicate that details of the business are to be found electronically, on the Internet". See further (but in a trade mark context) Kenny J in Sports Warehouse v Fry Consulting Pty Ltd (2010) 186 FCR 519 at [155]. 129 When considering whether the use of "www.solarhut.com.au" was deceptively similar to "SOLAHART" in relation to an allegation of trade mark infringement, Perram J in Solahart Industries Pty Ltd v Solar Shop Pty Ltd (2011) 281 ALR 544 said at [49]: In the context of domain names, one should put to one side in the analysis the various accoutrement of domain names such as "www" and ".com.au" to which consumers do not play any substantive attention. 130 The tension created by ".com.au" being both essential to the term "realestate.com.au" being recognisable as a name or brand but at the same time subsidiary (because substantive attention will not ordinarily be given to it), means that the term is not readily recognisable as a name. In the absence of prior familiarity, it is only faintly recognisable as an intended name for a business or brand. The act of recognition requires a focus and attention not required when a distinctive word or phrase is recognised as a name or brand. The additional scrutiny given to a name not readily recognisable as such, means that in a comparative exercise with a very similar term, including an exercise based on recollection, small differences are more likely to be noticed. 131 Thirdly, when taken as a whole "realestate.com.au" is, at least in part, descriptive of a real estate portal. REA conceded that the name was partly descriptive of the services offered at the website. The use of a name which is descriptive of REA's activities was as Stephen J said in Hornsby "no doubt convenient" (at 229). REA's evidence acknowledged that its name had given its business a natural prominence. That is entirely unsurprising in relation to a business which relies heavily on search engine searches by consumers to bring business to its door and where the search term "real estate" will most often be the keyword phrase used in a search. 132 The natural applicability of the phrase "real estate" in combination with a suffix, to indicate a real estate portal, is exemplified by its extensive use by real estate portals in both trading and domain names. Two of the three most popular Australian portals, "realestate.com.au" and "realestateview.com.au", have those features. Beyond "realestateview.com.au", the evidence shows there are many property portals with domain names which are only slightly different to "realestate.com.au", including: therealestate.com.au; realestateone.com.au; realestate1st.com.au; realestateworld.com.au; ingrealestate.com.au; realestateguide.com.au; realestatesource.com.au; westrealestate.com.au; abcrealestate.com.au; prime-realestate.com.au; loverealestate.com.au; bellrealestate.com.au; realestatewa.com.au; and realestateplus.com.au. 133 A keyword search for "realestate.com.au" in evidence shows that the term "real estate" appears some 50 times, and the term "realestate.com.au" appears some 38 times (twice with hyphens, twice without dots), as or in domain names shown in URL's in the first 100 results. Whilst that count involves some duplication of the same business because more than one search result appears for some businesses, it nevertheless demonstrates the extensive use of the descriptive terms in question on the very canvass which this case raises for consideration. 134 In this sphere of commerce, the phrase "real estate" in association with a top-level suffix is in common use and particularly so as a domain name or part thereof appearing on a search results page. That situation exists without apparent difficulty. Where common descriptors are utilised by businesses, cues as to the identity of a particular business are likely to be less pronounced and less differentiated from those of many of their competitors. That circumstance will be recognised by consumers. Consumers can be expected to be more vigilant when searching for a particular online real estate business and small differences are less likely to lead to error. 135 The term "realcommercial.com.au" does not suffer the same limitations as does "realestate.com.au". It contains a distinctive dominant term which, in and of itself, connotes the existence of a source business or brand. "Real Commercial" will be assumed to be the name or at least the essence of the name of the entity behind the term. Small differences to "Real Commercial" in the name of a rival, are unlikely to avoid misrepresentations of an association or affiliation.

Were members of the class likely to have been misled or deceived by the use of "realestate1.com.au"? 136 REA's complaint, as expounded at trial, was narrower than that made by its pleadings in a number of respects. As this part of the case was run at trial, no serious attempt was made to establish that when clothed in its logo, colour and other brand elements, the term "realestate1.com.au" was insufficiently distinct from the name, logo and other brand elements of realestate.com.au to be misleading or deceptive. REA did not press that the use of the impugned domain name on Real Estate 1's residential property portal was misleading or deceptive. Even if that were established, there was no connection made in the evidence between that conduct and any damage to REA, or why the relief sought by REA is otherwise justified. 137 The focus of REA's case was the use of "realestate1.com.au" in a URL on search results pages. That was the doorway at which traffic intending to go to REA's realestate.com.au portal could be diverted to Real Estate 1's residential property portal. As expounded at trial, REA's case was that such a diversion had likely occurred as the result of specific conduct and in specific circumstances. REA's contention was that a consumer familiar with one of its websites and intending to gain access through a search engine would be diverted to Real Estate 1's residential portal because the consumer would either: (i) see the additional "1" in "realestate1.com.au" in the URL of an organic search result or a sponsored link and assume that the website concerned was affiliated or connected with REA; or (ii) not see the additional "1" in "realestate1.com.au" in the URL of an organic search result or a sponsored link and click onto a Real Estate 1 search result believing that the consumer was accessing the realestate.com.au portal. 138 I will consider first the contention that the use of realestate1.com.au suggests an affiliation with REA's residential portal. 139 I have found earlier that a small difference in the name of a rival of realestate.com.au will likely suffice to avoid consumers being deceived. REA accepted that whether a sufficient difference exists in the name of a rival is a question of degree. REA pointed, by way of example, to the name of one of its primary rivals "realestateview.com.au" as being sufficiently differentiated from "realestate.com.au". It contended however that "realestate1.com.au" was on the wrong side of the line. Real Estate 1 argued that the additional "1" is sufficiently prominent and is a signal of disassociation rather than of association for anyone familiar with realestate.com.au. 140 Much of REA's focus was placed on there being only one keystroke difference between the two terms. However, the current analysis presumes that the additional "1" has been noticed. Although noticed as a single digit figure, the additional "1" will be understood to represent the word "one". When the term "realestate1.com.au" is read, whether read out aloud or not, the figure "1" will be of no less prominence to the mind of the reader than if the word "one" had appeared instead. The word "one" seems to me to be not significantly less distinctive than "view" when read between "real estate" and ".com.au". An additional single keystroke which does not connote a word would be quite different. For instance, the addition of an "s" to create "realestates.com.au" may not amount to a sufficient differentiation. However, in the case of the additional "1", its limited physical presence is not a true reflection of its prominence. 141 There are other differences between "realestate" and "realestate1" which belie their physical similarities. One is a concocted word the other is not. The term "realestate1" is recognisable as a name or brand, "realestate" is not. In a context where ".com.au" may be given little attention, those differences are significant. 142 Whether the figure "1" when attached to another term will be perceived as an indication of association, or alternatively of disassociation, is, I think, dependent on placement and context. The figure "1" like the term "plus" may be utilised to express an extension of, or a derivative of, a brand or a name. Such a device will usually be placed at the end of a name or brand and not in the middle. It is a device which could suggest an association when placed next to an identifiable name but does not have the same effect when placed next to a descriptive term. That point can be made by reference to two of the names in play in this case. To my mind, the term "realcommercial1.com.au" suggests an association with the entity or brand "realcommercial". The term "realestate1.com.au" far less so, because "realestate", for the reasons already explained, does not suggest an entity or brand. 143 Mr Coker founded his conclusion that some consumers may assume an affiliation or association on two bases. He raised the possibility that a consumer would perceive realestate1.com.au to be a "sub-domain or mirror site" of the realestate.com.au website. This suggestion was based upon the existence of an internet convention in which a numeral is added to the "www" element of the URL to denote a mirror site (eg www1.brand.com). Mr Coker suggested that a consumer who does not have knowledge, or who has insufficient knowledge of how URL's are structured would be more likely to confuse the URL to be a sub-domain or mirror site of realestate.com.au, than an individual who had sufficient knowledge of how URL's are conventionally structured. Mr Coker's evidence raised this prospect as a mere possibility and, in my view, the possibility is more hypothetical than real. I would not ascribe to the 'ordinary' or 'reasonable' persons in the class an appreciation of the existence of the technical convention relating to sub-domain or mirror sites. 144 Secondly, Mr Coker thought there was a possibility that a consumer would assume that realestate1.com.au was a variant of and thus affiliated with realestate.com.au if the consumer had a poor recall of the name "realestate.com.au". On the other hand, Mr Coker also thought there was a possibility that a consumer with a poor recall who was confronted with a page or list in which variants of "realestate.com.au" were commonplace, would see each as different, separate and unassociated. 145 I accept that an erroneous assumption of affiliation as a consequence of poor recall is possible. The extent to which it would happen or likely happen was not the subject of Mr Coker's evidence or any other evidence. However, even assuming occurrences of that kind to be significant, I do not regard any erroneous assumptions of affiliation based upon poor recall as necessarily flowing from the conduct of Real Estate 1. Lack of recall will at least in part be a function of the indistinctive name chosen by REA. That kind of confusion is the kind that Stephen J said "was not anything to which s 52 is directed". 146 For the reasons earlier identified, consumers will be more vigilant in the context of a market where close variants of "realestate.com.au" are common on a search results page. Consumers will likely recognise that the use of "realestate" as a central feature in many domain names on the page to be a consequence of the natural applicability of the term to a real estate portal. Small differences in description from the descriptor of the article being traded are likely to be perceived as marks of distinction as between many traders rather than an indication that one trader is utilising a variety of descriptors for its product or brand. I think the better conclusion is that a consumer with ordinary recall specifically looking for "realestate.com.au" on a search page who sees the "1", will not be misled or deceived into assuming an affiliation between the portal being searched for and "realestate1.com.au". 147 I turn then to consider the second kind of deception alleged by REA based upon the consumer failing to notice the additional "1". This was REA's primary contention upon which most of its submissions were based. 148 REA tendered a small number of examples in which the term "realestate1.com.au" appeared embedded in the URL element of an organic search result. Other examples of organic search results for Real Estate 1's residential property portal were tendered by Real Estate 1. There were no examples tendered of searches in which sponsored links appeared for Real Estate 1's residential property portal, however the evidence included the specific content of sponsored links and allows for the form and content of those sponsored links to be reconstructed. In the period 1 April 2009 to 31 January 2012, Real Estate 1 ran 10 different search engine marketing campaigns. A report from Google AdWords provides the text of the various sponsored links used by Real Estate 1 in those campaigns. That report was annexure "ROM-23" to the affidavit of Ryan McAvoy ("the ROM-23 report"). 149 It is helpful to set out some examples from the evidence of search results for Real Estate 1's residential property portal. The following is an example of an organic result: Real Estate Listings - Search & Find Property for Sale & Rent… www.realestate1.com.au/ Search and find all types of real estate in Australia. Use our unique search by street, filter by agency, agent, suburb or region for your next home, land package… ("The first example") 150 The following are examples of sponsored links reconstructed from the ROM-23 report: Real Estate 1 www.realestate1.com.au/real-estate Looking for real estate in VIC. Look no further than realEstate1 ("The second example") Find Real Estate www.realestate1.com.au/ Our aim is to make searching for real estate pure, simple and fast! ("The third example") Echuca Real Estate www.realestate1.com.au/Melbourne Rent/Buy/Sell real estate in Echuca. Contact realEstate1. ("The fourth example") 151 What is immediately apparent from the above examples, is that the content of search results deployed by Real Estate 1 differed. In some, the term "Real Estate 1" was used in the heading. In others "realestate1" was used only in the URL. In yet others the use of the term in the URL was accompanied by a reference to "Real Estate 1" in either the heading alone, the descriptive text alone or in both. 152 REA's contentions, and in large part the evidence of Mr Coker, focussed upon a search result in which the only cue to the identity of the source business to which the result related, was the "realestate1.com.au" cue in the URL. The first example was referred to as a "typical" search result. The evidence does not support that assertion. 153 REA's contention was essentially that whenever presented in a URL on a search results page before a consumer searching for the realestate.com.au website, the term "realestate1.com.au" was a representation that was demonstrably false. 154 Some words spoken or written may be demonstrably false, without a consideration of the context or surrounding circumstances in which they were stated. However ordinarily, in order to characterise the nature of the representation in question, it is necessary to take into account the context or surrounding circumstances in which the words were presented: ConAgra Inc v McCain Foods (Aust) Pty Ltd (1992) 33 FCR 302 at 380 (French J). As French J there said: "Where the name and get-up of a product are in issue, the question for the purposes of s 52 is whether they are misleading or deceptive in the circumstances." 155 It is well established that in considering whether consumers are likely to be misled or deceived, "whether or not conduct amounts to a misrepresentation is a question of fact to be decided by considering what is said and done against the background of all surrounding circumstances": Taco Bell at 202 (Deane and Fitzgerald JJ). In Campomar, the High Court referred to that observation with approval and emphasised that, "[t]he matter should not be considered in the abstract" and that, "[r]egard must be had to the circumstances of the particular case" (at [99] and [100]). 156 One of the obvious circumstances that must be considered in relation to an impugned representation is what else was said or done on the same occasion. A search result, as the examples above show, is a short three or four line message. Unless a basis exists not to do so, the whole of the message should ordinarily be taken into account when characterising the representation that it conveys on a particular subject matter. In this case, the relevant subject matter is the identity of the portal business to which the search result relates. Cues as to the identity of a portal business to which a search result relates may be found in the heading, URL or descriptive text of a search result. Additionally, beyond the search result itself, there may be other contextual circumstances on the particular search results page that may be relevant and which must be considered in determining whether readers were likely to be misled. 157 There was no issue that in a search conducted by a consumer looking for realestate.com.au by name, an organic result for that website would consistently appear at the top of the organic search results shown on the first results page. Mr Hudson's evidence was that part of his job for REA was to monitor how different websites rank on a search results page when the term "realestate.com.au" is used as the search keyword. His evidence was that invariably in relation to such a search, a search result for Real Estate 1's residential property portal would not appear before page 5 and suggested that it commonly appeared well below that page. 158 I have already referred at [133] to an example of a search conducted using the keyword "realestate.com.au". Whilst a result for the realestate.com.au website appeared as the first organic result, there is no result at all for Real Estate 1 in the 100 organic results shown. Another example in the evidence, this time for a search for the keyword term "real estate", shows a result for the realestate.com.au website as the first organic search result. Across the 100 search results shown in the example, the term "realestate" appears embedded in 28 URLs. The term "realestate.com.au" appears embedded in ten URLs. A search result for Real Estate 1's residential property portal does not appear until the bottom of page 8, some 80 entries later than that for REA's residential real estate portal. 159 There are two examples in the evidence in which an organic search result for Real Estate 1's residential property portal appeared on the first page of the search results. It was not suggested by REA that those two examples should be regarded as likely to have resulted from a consumer searching specifically for the realestate.com.au website. The first relates to a search with the keyword search term "houses for sale Melbourne". In that example results for the realestate.com.au website appear as the first three results. A result for the "realestateview.com.au" website appears as the sixth result and Real Estate 1 has the eighth result. The second example relates to a search using the keyword search term "property for sale". Results for the realestate.com.au website appear as the first and second organic results. A result for the realestateview.com.au website is the fourth result and a result for Real Estate 1 appears as the eighth result on the page. 160 It was not suggested by REA that a consumer familiar with the names of its residential property portal will always be misled, when reading "realestate1.com.au" in the URL element of a search result, into thinking that the website to which the search result refers is the realestate.com.au website or an associated website. If a consumer sees the additional "1" in the URL, then REA's evidence given through Mr Coker acknowledges that where the consumer has good recognition of the name "realestate.com.au", the consumer will likely assume that the URL is the address of a different and unrelated website. That must be so, as the 'ordinary' or 'reasonable' persons in the class will know that URLs are unique. 161 REA's contention was that deception was likely because the similarity between "realestate1.com.au" and "realestate.com.au" would cause many consumers who were searching for "realestate.com.au" to miss the additional "1" and thus assume that the URL they were looking at was REA's residential portal. 162 To evaluate that contention, it is necessary to consider the manner in which a consumer who knows the realestate.com.au portal by name and is searching for that portal specifically, is likely to deal with the search results displayed. The 'ordinary' or 'reasonable' consumer I have in mind is a consumer who has sufficient brand recollection of the name "realestate.com.au" to be able to recall that name. 163 I do not consider that REA can make out its case in relation to consumers who have an imperfect recollection of the name of the portal the consumer desires to visit. In the context in question, a consumer who is searching for a portal recalled by the consumer as "real estate something" and ends up at Real Estate 1's website has not been misled or deceived. As I said earlier, if such persons have been diverted, the fault may lie with the adoption by REA of a name insufficiently distinctive to have facilitated its recall. A sufficient nexus between any conduct of Real Estate 1 and any misconception by the consumer is not established in that circumstance: Campomar at [98]. 164 Given that the relevant consumer will know the website by name, he or she will most likely type "realestate.com.au" or "www.realestate.com.au" into the search bar when initiating the search. In the expectation that organic search results are primarily ordered by reference to their relevance to the keyword searched for, such a consumer will head for the first organic result because that is where the consumer will expect the clickable link to the desired website to be found. On the evidence before me, a consumer who typed in "realestate.com.au" or a variant thereof will invariably have found an organic result for the realestate.com.au website at or near the top of the first page. 165 In this process, the consumer is looking for an identity cue to confirm the website searched for. A natural place to look for that cue is the URL. As the consumer expects to see "realestate.com.au" embedded in the URL at or near the top of the organic results, I accept that it is possible that the consumer will scan the URL without noticing an additional "1" between "real estate" and ".com.au", if contrary to the consumer's expectation, "realestate1.com.au" is embedded in the URL at or near the first organic result. 166 In those circumstances, I think that a not insubstantial number of consumers may well miss the additional "1" in the URL. In a URL both "realestate.com.au" and "realestate1.com.au" appear in the same font and in small print. The distinction between the two terms is confined to the additional "1". Whilst I consider that most 'ordinary' or 'reasonable' consumers involved in the scenario here contemplated would see the difference, I accept that many would not. 167 However, there is no evidence whatsoever to suggest that in relation to a search of the kind here being considered, an organic result for Real Estate 1's residential property portal has ever appeared at or near the top of the organic results or indeed on the first five pages. On the evidence, I am not satisfied that a not insignificant number of consumers have or were likely to have been misled or deceived in the manner here considered. 168 My consideration of this issue has only so far dealt with the issue of deception at the top of the organic results. Any deception of that kind is at least in part, a product of the expectation of the consumer that realestate.com.au will be found at or about the first result. Once the attention of the 'ordinary' and 'reasonable' consumer moves away from the first result and to the rest of the search results page, that expectation will no longer be in play and the 'ordinary' or 'reasonable' consumer may be expected to be vigilant to notice small differences on a page riddled with references to "realestate" and "realestate.com.au". 169 Whilst evidence of actual consumer confusion can be of great weight (Australian Woollen Mills Ltd v FS Walton & Co Ltd (1937) 58 CLR 641 at 658 (Dixon and McTiernan JJ)), none was directly called by REA. The circumstantial evidence that REA relied upon to suggest that in the order of 2,000 consumers who had specifically searched for its website were deceived into choosing an organic search result for Real Estate 1 because they had missed the additional "1" in the URL, was not persuasive. REA relied on two reports which evidenced the keyword search term utilised by consumers who had visited the realestate1.com.au website having got there by clicking on an organic search result for that site. On about 2,000 occasions the keywords utilised were "realestate.com.au" or a close variant thereof. REA also relied upon other evidence which showed that a high proportion of visitors to the realestate1.com.au website left the site after only viewing the first page and that the most popular site visited thereafter was the realestate.com.au website. That evidence was said to demonstrate the realisation of error and its immediate correction. 170 The evidence of the consumer behaviour in question was open to a number of explanations not necessarily consistent with deception. The evidence of time spent by such consumers on the realestate1.com.au site did not consistently suggest the realisation of error and immediate correction. The evidence that the most popular next visited site was realestate.com.au does not lead to the conclusion REA has asked me to draw. That evidence was general and not specific to those consumers searching for REA's website. It is unsurprising, given the overwhelming popularity of REA's website, that the next most popular visited site was realestate.com.au. That pattern would likely apply in relation to all residential property portals. 171 There are, however, two overwhelming difficulties in the Court drawing the inference that REA invites the Court to draw from the circumstantial evidence which REA suggested identified some 2000 deceptions. The first is that in relation to the same period to which the evidence relied on by REA relates, Mr Hudson's evidence was that an organic result for Real Estate 1 did not appear before page 5, when a search was conducted using the keyword term "realestate.com.au". I have great difficulty in inferring that a searcher looking for realestate.com.au who travelled to page 5 or later before coming across a result for Real Estate 1, would have been deceived in the manner REA suggests. Secondly, most of the 2,000 instances relied upon occurred in three months of the 22 month period covered by the reports. The trend over the 22 month period shows relatively few visits per month to Real Estate 1's residential portal said to be the result of deception. Why there was an extraordinary spike of allegedly deceived visitors in January to March of 2010 was not explained other than a suggestion in REA's submissions that Real Estate 1 was active with search engine optimisation in the period in question. No evidence was pointed to in support of that bare assertion. In any event, the assertion is inconsistent with Mr Hudson's evidence. If Real Estate 1 was particularly active during those three months, that activity was not reflected in the observations made by Mr Hudson. 172 It is also necessary that I express my view on the issue of causation. Even if I am wrong as to the likelihood of deception, in the context in question (ie the use of "realestate1.com.au" as an internet address), I would not be satisfied that a sufficient nexus is established between Real Estate 1's conduct and any deception. It needs to be remembered that REA makes no complaint about the use of "realestate1" alone. It only complains about the use of that term when ".com.au" is attached to it. That the TPA may be used by one competitor to deny another the capacity to attach the common suffix ".com.au" to its trading name (in this case, its registered trade mark) when that name and the suffix is used as an internet address, is a troubling thought. Stephen J in Hornsby held at 231 that to deny to traders the use of a locality prefix in association with a wholly descriptive phrase like "art gallery" would involve treating s 52 of the TPA as "the source of a newly created monopoly heretofore unknown to the law and likely to deter new entrants into a field which ought to be open to legitimate competition". His Honour found it difficult to contemplate any less likely legislative source for such a consequence. 173 In my view, a suffix like ".com.au" is, relevantly, of no different character to a locality prefix of the kind which Stephen J had in mind. It seems to me therefore that where a consumer mistakenly associates ".com.au" when appearing in an internet address as a part cue for "realestate.com.au" the mistake is attributable to the adoption by REA of the most common element of an internet address as a part of its trading name in combination with a descriptive second-level domain. Because ".com.au" is descriptive, it is equally applicable to any business of a like kind: Hornsby at [229] (Stephen J) and Cadbury Schweppes Pty Ltd v Pub Squash Co Pty Ltd [1980] 2 NSWLR 851 at 857 (the Court). Any loss of custom is the risk that a company takes when it chooses to incorporate into its own name words that are merely descriptive: British Vacuum Cleaner Company at 328 (Parker J). It is "the price to be paid" for the advantage of a descriptive trade name: Hornsby at 229 (Stephen J); Cadbury Schweppes at 857 (the Court). 174 The difficulty for REA raised by the evidence that an organic search result for Real Estate 1 appeared at least 5 pages beyond where a result for realestate.com.au first appeared, did not extend to sponsored links. There was evidence that Real Estate 1 bid for sponsored links and that in the period 1 April 2009 to 31 January 2012 over 400,000 clicks to its realestate1.com.au website were generated from sponsored links with Google. The ROM-23 report showed the content of those sponsored links. Although there was no direct evidence as to their positioning or likely positioning on a search results page, the general evidence as to the positioning of sponsored links suggested that mostly, sponsored links appear on the first search results page either above or to the right of the first appearing organic results. 175 REA contended that on at least 10,000 occasions between 1 April 2009 and 31 January 2012, a consumer who searched for the realestate.com.au website by name in a Google search engine, was misled into triggering a sponsored link for the realestate1.com.au website. REA relied upon two AdWords reports (Annexures ROM 11.1 and ROM 11.2 to the Affidavit of Ryan McAvoy) which recorded visits made to Real Estate 1's residential property portal after a consumer clicked onto a sponsored link for that portal. The reports were formatted by reference to various keywords used by searchers. Of the more than 14,000 visits recorded, at least 10,000 related to visits generated by keyword searches suggestive of the domain name "realestate.com.au". Although I have some reservation, I am prepared to infer that a substantial number of the 10,000 consumers here in question held an intent, at the time they initiated the search, to access the realestate.com.au website. 176 REA contended that the consumers in question must have had their attention drawn to the URL displayed in Real Estate 1's sponsored link. Its contention is that having seen the URL containing the term "realestate1.com.au" and having failed to notice the "1", the consumer was misled into believing that by clicking the heading he or she would arrive at the realestate.com.au website. 177 I do not accept that the evidence establishes that REA's contention is the likely explanation for why these consumers arrived at Real Estate 1's website. 178 As I have previously said, an 'ordinary' or 'reasonable' consumer who had typed "realestate.com.au" into a search bar in order to bring up a hyperlink to that site, would have expected the hyperlink to be shown at the top of the organic results. Such a person would understand the difference between an organic result and a sponsored link. Like all 'ordinary' or 'reasonable' users of search engines, such a person would regard an organic search result to be a more reliable source for the information searched for than a sponsored link. Such a person has no reason to look to a URL in a sponsored link for an identity cue. 179 However, advertising is designed to draw the interest and attention of people who had no prior intent to engage with it and that purpose is often achieved. The evidence demonstrates that the attention of the consumers here in question was in fact drawn to Real Estate 1's sponsored link. The real question is - what was it in the sponsored links that drew their attention? REA's contention is based on a consumer who has not carefully read the search page but who has scanned it and whose attention has been drawn to Real Estate 1's URL in a sponsored link. The eye of a consumer looking for an organic result may well have scanned across a sponsored link. However, in a scan of that kind, the consumer's eye is far more likely to be drawn to the heading of a sponsored link rather than a domain name embedded in a URL. Not only is the heading more visible but it is also the place where the advertiser is able to ply its trade. 180 People looking for information about real estate will likely have a particular geographic location in mind. It is plausible that a consumer who has searched for a particular property portal may be diverted from that path by a reference in a sponsored link which happens to identify the very geographical location the searcher had in mind. Not only is such an explanation plausible in a general sense, it is an explanation supported by the evidence upon which REA relied. 181 In the period 1 April 2009 to 31 January 2012, Real Estate 1 ran ten different search engine marketing campaigns. The ROM-23 report shows the text of the various sponsored links used by Real Estate 1 for those campaigns. The further AdWords report shows the keywords bid for by Real Estate 1 for each sponsored link in the various campaigns and the number of clicks a successful bid generated ("the ROM-20 report"). 182 Of the just over 420,000 clicks generated by the ten campaigns, one campaign ("the Real Estate Locations Campaign") was, in relative terms, an outstanding success. The Real Estate Locations Campaign generated over 384,000 clicks. The campaign was designed to target searchers looking for property in one of some 1,800 suburbs or towns in Victoria which Real Estate 1 decided to target. Putting some minor exceptions to one side, the sponsored links for the campaign were nearly all in the same form apart from the named location which varied from one sponsored link to the next. The sponsored link for Echuca shown at [151] as the fourth example was typical of the sponsored links used in this campaign. 183 To generate sponsored links for the Real Estate Locations Campaign, Real Estate 1 bid for a wide range of keywords. Many of the keywords bid for included a particular location for instance - "Croydon real estate". However, Real Estate 1 also bid for keywords suggestive of the term "realestate.com.au". Most of those bids were for broad match keywords, although there were some for phrase match keywords. A successful broad match bid for keywords containing a combination of "real", "estate", "com" and "au" would be triggered by a search for the keywords "realestate.com.au" or variants thereof such as "real estate com au". 184 Keywords of that kind (ie those capable of being triggered by a keyword search for "realestate.com.au" or like terms), that were bid for by Real Estate 1 as part of the Real Estate Locations Campaign, generated over 14,000 clicks. Each of those successful sponsored links contained text like that contained in the Echuca example shown as the fourth example. The most successful sponsored link was for the Melbourne suburb of Research. That generated 11,865 clicks having been triggered by Real Estate 1's broad match keyword bid for "realestate au". 185 The evidence is supportive of the fact that the sponsored links used in the Real Estate Locations Campaign were displayed to many if not most of the body of searchers that REA suggests were misled. Given that to access the realestate1.com.au website each of the consumers in question would have had to place the cursor of their computer on a clickable heading like "Echuca Real Estate", it is far more likely that if those consumers were diverted they were diverted by a reference in a heading of a sponsored link to a particular geographical locality. 186 Perhaps the best example in the evidence of the kind of diversion that likely occurred relates to the single searcher whose keyword search was "advertise on real estate com au". That search can be traced through the AdWords reports. It triggered a Real Estate Locations Campaign sponsored link for "Echuca" which is the fourth example. The link was generated by Real Estate 1's phrase match bid for the keywords "advertise on real estate com au". I have little doubt that the searcher here in question was headed for the realestate.com.au website searching for a specific page which would provide information on how to advertise property on that website. Despite the searcher's intent to go to a highly specific webpage, the searcher must have placed his or her cursor on, and clicked onto "Echuca Real Estate". Unless the click occurred unintentionally, it is highly unlikely that it occurred without the searcher having altered the intent that he or she held when the search was commenced. 187 I accept that it is possible that such a searcher may have been diverted to a Real Estate Locations Campaign sponsored link by the heading and then clicked onto that link thinking that he or she was being taken to realestate.com.au's website because of the searcher's mis-reading of the URL. However the evidence in the ROM-23 report shows that all but a tiny minority of the sponsored links used in the Real Estate Locations Campaign contained text that referred to Real Estate 1 by name in the descriptive text. The descriptive text in the sponsored link typically ended with the following words - "Contact realEstate1"; or "search realEstate1"; or "Visit Real Estate 1 now". As I have said, it was not disputed that the name "Real Estate 1" is sufficiently differentiated from "realestate.com.au". An 'ordinary' or 'reasonable' member of the class familiar with realestate.com.au by name and who had read the text of the sponsored link is not likely to have been misled or deceived into believing it was an advertisement for realestate.com.au. Whilst its possible some may not have read the descriptive text, the numbers involved are not likely to be significant. In any event, for the reasons already given, in my view an erroneous assumption based upon Real Estate 1's use of ".com.au" in a URL does not have a sufficient nexus with any deception caused by Real Estate 1. 188 There is evidence that in the less successful campaigns run by Real Estate 1, some sponsored links were used in which the only identity cue given was in the domain name in the URL. Again, I do not consider that REA has established a sufficient nexus between the use by Real Estate 1 of its domain name in an internet address and any deception. In case I am wrong as to the sufficiency of a nexus, I should add that in relation to sponsored links of the kind to which I now refer, REA did not identify any evidence which established a connection, or a likely connection, between the clicks made on the sponsored links concerned, to consumers familiar with realestate.com.au by name and searching for the realestate.com.au website. 189 Finally, there is also a further category of visits to Real Estate 1's sponsored links (shown at C1938 the ROM-20 report) which the evidence there given suggests may have been displayed on search results pages generated by a searcher searching for the realestate.com.au website. However, the evidence does not provide the content of those sponsored links. It is therefore not possible to consider the context in which the use of "realestate1.com.au" in the URL occurred. 190 On the available evidence of Real Estate 1's organic results and sponsored link activities where "realestate1.com.au" was used in a URL on a search results page, I am not satisfied that REA has established that a not insubstantial number of 'ordinary' or 'reasonable' persons in the class have or were likely to have been misled or deceived. I turn now to explain two matters relevant to those conclusions which require further exposition.

A not insubstantial number of consumers 191 On the question of whether a not insubstantial number of consumers were likely to be misled by the impugned conduct of Real Estate 1, it was contended by Real Estate 1 that the number involved should be evaluated as a proportion of the total number of visits to the realestate.com.au website. Thus it contended that if consumers were misled, as a proportion of the total visits made to the realestate.com.au site, the proportion was infinitesimal because the evidence suggests that some 50 million visits per year are made to the realestate.com.au website via organic or paid search results. I have not adopted that approach. The numerical significance of consumers who may have been misled is to be assessed by reference to their number. If that were not so, large numbers of consumers would gain no protection against misleading or deceptive conduct simply because, in proportional terms, they represent a small fraction of a very large class. 192 However in determining whether I could be satisfied that a not insubstantial number of consumers within the class were or were likely to be misled, I have not included in that assessment consumers who would not have been deceived other than in a transitory and insignificant manner. 193 As French, Tamberlin and Rares JJ said in Knight v Beyond Properties Pty Ltd (2007) 242 ALR 586 at [58]: When characterising a course of conduct as misleading or deceptive, the practical consequences and effect which the conduct is likely to have must be taken into account. 194 Their Honours had earlier observed at [54] by reference to the judgment of French J in ConAgra that: A transitory or ephemeral impression, if misleading, but which is immediately dispelled, may, depending on the circumstances, be of no commercial significance and therefore not misleading conduct within s 52 of the TPA. See further: Lumley Life at 324 (Lockhart J) and Bing at [48] (Kenny J, with whom Greenwood and Logan JJ agreed). 195 A consumer using a search engine intent on accessing the realestate.com.au website, who had missed the additional "1" in "realestate1.com.au" and clicked the clickable heading of the search result, would have been taken immediately to Real Estate 1's residential portal. Such a person may or may not have had prior exposure to the realestate.com.au website. If the person had prior exposure it is highly likely that the person would immediately recognise that the wrong website had been accessed. The two websites were very different in terms of branding and user experience. A person with prior exposure to the realestate.com.au website would not likely have mistaken Real Estate 1's product for that of REA and would likely have returned to the search results page to find the correct hyperlink. 196 The situation is not dissimilar to that considered in Knight where a consumer who may have been misled by the name of a TV program would have had the misleading impression immediately dispelled upon actually watching the show. In Knight, the Full Court said at [58]: In this case, even if one assumes that a person consults a television guide and momentarily apprehends that there may be some association between Mr Knight's Mythbusters Books and the Mythbusters TV Show, but on closer inspection is immediately disabused of that false or confused impression, then it is open to find that the respondents' conduct is not misleading or deceptive. On the other hand, if the conduct leads that person to take action of a substantive nature or causes some significant detriment or financial loss, then the position may be different, and one may characterise the conduct in a different way. The descriptions "misleading" or "deceptive" in relation to the conduct are indicative of the importance of looking to the impact of that conduct. Put another way, if the conduct only produces the effect of misleading a person for a transient period and to an insubstantial extent, as it did in this case, then this indicates that the conduct is not likely to be misleading or deceptive in any commercially significant sense. 197 That passage, as I understand it, looks at the issue of commercial significance from the perspective of the consumer and not from the perspective of the trader who may have lost a customer. That must be so, given that the significance of a respondent's conduct goes to the issue of whether the consumer has been misled, rather than the commercial consequence for another of any misleading conduct. 198 In any event, REA's point that there was a commercial consequence for it, was not made out. The only commercial consequence for REA which was pointed to, where a consumer was momentarily diverted to Real Estate 1, was said to be the benefit gained by Real Estate 1 of an additional visitor to its website. I accept that the greater the traffic to a portal, the more likely it is that advertising will be attracted to it instead of its competitors. However there is no necessary correlation between the gain by Real Estate 1 of additional visitors and the loss of advertising to REA. Beyond the bare suggestion, REA made no attempt to establish any such correlation. Given the disparity of the size and traffic volumes involved, I doubt that the existence of Real Estate 1, let alone a marginal increase in traffic to its website, would be appreciated by those who advertise with REA. 199 It is possible that some consumers without prior experience of the realetstate.com.au website, but who knew it by name (for instance where the site was recommended by a friend), will not have appreciated that the Real Estate 1 website accessed was not a REA site. While I recognise it as a possibility, given the improbability of the sequence of events required for such a person to have been diverted to a Real Estate 1 website, I consider the numbers involved were not likely to be material.

Real Estate 1's intent 200 In contending that a finding should be made that Real Estate 1's conduct was misleading or deceptive (and constituted passing off), REA asserted that the conduct was engaged in with intent to take advantage of the goodwill associated with REA's websites. Whilst that was a matter which I took into account when reaching my conclusion that REA has not established misleading or deceptive conduct (or passing off) in relation to the impugned use of "realestate1.com.au", it was not a matter that affected that conclusion for the reasons which follow. 201 There is no requirement for REA to prove any subjective intention by Real Estate 1 to mislead or deceive or to pass off its portals as that of REA's. However, a trader's intent may say something about the likelihood that its conduct has or is likely to be misleading. That is because "the courts have long recognised that trader's best know their trade": Telmak Telproducts (Aust) Pty Ltd v Coles Myer Limited (1989) 89 ALR 48 at 69 (Wilcox and Einfeld JJ). In Australian Woollen Mills at 657, Dixon and McTiernan JJ reinforced the view that intent may have an evidential value. Their Honours said: The rule that if a mark or get-up for goods is adopted for the purpose of appropriating part of the trade or reputation of a rival, it should be presumed to be fitted for the purpose and therefore likely to deceive or confuse, no doubt, is as just in principle as it is wholesome in tendency. In a question how possible or prospective buyers will be impressed by a given picture, word or appearance, the instinct and judgment of traders is not to be lightly rejected, and when a dishonest trader fashions an implement or weapon for the purpose of misleading potential customers he at least provides a reliable and expert opinion on the question whether what he has done is in fact likely to deceive. 202 It is well established that where there is a finding of intention to deceive, a court may more readily infer that the intention has been or, in all probability, will be effective: Campomar at [33]; and see Bodum at [216]. 203 By reference to that line of authority, REA sought to establish that Real Estate 1 held an intention to deceive consumers and that a finding of such an intent tends to establish that the conduct was misleading or deceptive. 204 I accept that a 'rebranding' exercise that commenced in about March 2009 and which continued at least through to the commencement of this proceeding, involved a deliberate strategy by Real Estate 1 to have its brands closely resemble the brands utilised by REA for REA's property portals. That strategy involved the rebranding of the logo for Real Estate 1's residential portal to "realEstate1.com.au" in March 2009 and in August 2009, the rebranding of the logo for Real Estate 1's commercial portal to "realCommercial1.com.au". In each case the change involved a new domain name to reflect the logo. Mr Luff also organised for the registration of the name "realholiday1.com.au" and in 2010 announced a "realholiday1" portal as a forthcoming addition to Real Estate 1's business. The similarity between Real Estate 1's family of brands and that of REA could hardly have been coincidental. 205 There were good commercial reasons for Real Estate 1 to replace ".net.au" with the far more common ".com.au" in the internet addresses for its portals when that opportunity became available. There were also good commercial reasons to adopt the term "realestate" as part of its trading name, given the natural prominence the name gives to a real estate portal. Whilst those commercial advantages played a part, I have no doubt that the branding adopted by Real Estate 1 was also influenced by the presence of REA as the leading property portal business in the Australian market. 206 Part of Mr Luff's strategy for the Real Estate 1 businesses was an exit plan in which he hoped Real Estate 1 would be taken over by REA. I accept Mr Scaife's evidence to that effect. Mr Luff was well aware that the names deliberately chosen by Real Estate 1 for its residential portal were very close to those used by REA. That much is apparent from his comment to prospective investors in December 2010 that - "Our choice of name and brand was no accident. We wanted to capture market share in this space quickly. A little cheeky I know. But so be it. It has worked". 207 Given the relative size of REA's business, I do not consider it likely that Mr Luff's exit strategy involved a belief that as a significant commercial rival to REA, Real Estate 1 would be bought out by REA. It seems more likely to me that Mr Luff's exit strategy was based upon his view that REA would perceive Real Estate 1's branding as an irritant and may be persuaded to take it over to protect its brands. 208 In my view, Mr Luff was motivated to rebrand for a mix of reasons. Some commercial and some far less honourable. 209 He was certainly motivated to try to divert traffic from realestate.com.au and other real estate portals through the search engine marketing exercises in which Real Estate 1 was engaged. However, I very much doubt that such conduct is unusual as between on-line competing businesses and an intent to divert customers does not necessarily involve an intent to deceive. As Parker J said in the Vacuum Cleaner Case at 328, it is perfectly legal for a rival company to set up in the same street, two doors down. 210 The issue is not whether Mr Luff had questionable business practices or motives. The pertinent question is whether in rebranding Real Estate 1's residential portal, Mr Luff did so because he was, at least in part, of the view that using the domain name "realestate1.com.au" in a URL on a search results page would likely deceive consumers searching for realestate.com.au into believing it was that website. That is the specific question the Court is called upon to answer and it is the issue to which Mr Luff's intent may provide some evidential value. 211 It is likely that Mr Luff's rebranding exercise was done with the intention of pressuring REA to purchase the business at some later time. Mr Luff's comment that "we wanted to capture market share in this space quickly", is said by REA to substantiate an intent to profit on the coat tails of the market leader by adopting a similar name. The comment was explained by Mr Luff as based on his naive perception that the use of the name would result in realestate1.com.au ranking highly in online search results. Given his exit strategy, that explanation is not likely to be the whole truth. I am satisfied that Mr Luff was at least in part motivated to adopt realestate1.com.au as his brand because he wanted to take advantage of the reputation of realestate.com.au. That finding does not necessarily carry with it a finding that Mr Luff held the view that some consumers would be misled or deceived in the manner contended for by REA. 212 Even if I had made such a finding, I would not have given that finding significant weight. Mr Luff's knowledge and understanding of the market in which Real Estate 1 operates was unsophisticated and unreliable. He readily conceded his naivety. Neither he nor anyone associated with Real Estate 1 had any prior experience of or expertise in online trading or consumer behaviour on the internet. That "traders best know their trade" is not a rule of universal application. I do not think that any view held by Mr Luff that consumers would be deceived, in the specific circumstances which arise for determination in this case, provides me with any reliable evidentiary assistance. In Windsor Smith Pty Ltd v Dr Martens Australia Pty Ltd (2000) 49 IPR 286, Sundberg, Emmett and Hely JJ considered that the deliberate adoption by a trader of a rival's get-up in order to imply a connection with the rival, was "simply one piece of evidence to be assessed with such other evidence as may be adduced on the issue": at [33] and see NEC Corporation v Punch Video (S) Pte Ltd (2005) 67 IPR 17 at [37] (Branson J). It should also be borne in mind that conduct engaged in with intent may nevertheless have failed in its purpose: Cadbury Schweppes at 860-861 (the Court); S & I Publishing v Australian Surf Life Saver Pty Ltd (1998) 88 FCR 354 at 362.

Conclusion 213 For all those reasons I am not satisfied that by the use of the term "realestate.com.au" in a URL on a search results page, Real Estate 1 engaged in misleading or deceptive conduct in contravention of s 52 of the TPA. The conclusions I have reached in relation to the s 52 claim also require that the s 53 claims and the corresponding passing off claim be dismissed.

Were members of the class likely to have been misled or deceived by the use of "realcommercial1.com.au"? 214 At trial, REA tendered evidence which established consumer familiarity with the realcommercial.com.au portal. REA also tendered evidence which established that the term "realcommercial1.com.au" was used, at least until October 2010, as the trading and domain name for Real Estate 1's commercial property portal. 215 However, REA did not at trial expound a case to establish that when clothed in its logo, colour and other brand elements, the name "realcommercial1.com.au" was used in a manner likely to mislead or deceive. No case was pressed that the use of realcommercial1.com.au on Real Estate 1's commercial property portal was misleading or deceptive. 216 The use of "realcommercial1.com.au" on search results pages and how such use would likely mislead or deceive was not the subject of any evidence or specific submission. If such a case was being seriously pressed on the same general basis as that pressed in relation to the use of "realestate1.com.au" in URL's, I would have rejected it. 217 I would have accepted that, absent a proximate and prominent disclaimer to the effect that no association was intended, the use of "realcommercial1.com.au" in a URL on a search results page would likely have suggested an affiliation with the realcommercial.com.au portal. For reasons I have already expressed, the addition of the figure "1" to a distinctive word which is recognisable as a name is likely to suggest to many consumers that realcommercial1.com.au is affiliated with realcommercial.com.au. There was no evidence that close variants of the name "realcommercial.com.au" were a common feature of the market for commercial real estate portals. 218 I would have found a sufficient nexus between Real Estate 1's conduct and any such suggestion of affiliation. The necessary nexus is established because it was Real Estate 1's use of "realcommercial1" rather than the suffix ".com.au" which was responsible for any likely deception. 219 However, for largely the same reasons I have given in relation to the use of "realestate1.com.au" on a search results page (see paragraph [194] to [201] above), I would not have been satisfied that any likely deception would have been other than momentary. It is likely that any confusion would have been immediately dispelled by the consumer accessing the realcommercial1.com.au website. There was no evidence of the impact or likely impact of such confusion which would have permitted a finding that Real Estate 1's conduct was of any commercial significance. 220 In the circumstances, REA's ss 52 and 53 claims and its claim of passing off in relation to the use by Real Estate 1 of "realcommercial1.com.au" should be dismissed.

Trade Mark InfringemenT 221 REA claims that the following marks used by Real Estate 1 infringe REA's registered trade marks:

REA's Mark Real Estate 1's uses 222 REA makes no complaint about Real Estate 1's use of the mark "realestate1" but when that mark is used in conjunction with ".com.au" in relation to the websites of Real Estate 1, REA claims that such use infringes its registered trade marks. The impugned use includes use of the marks on Real Estate 1's websites, in its advertising and promotional activities and extends to use of the marks in the title of sponsored links and as a domain name address appearing in a URL. 223 In relation to the realcommercial.com.au mark, REA complains about Real Estate 1's use of both "realcommercial" alone and its use in combination with ".com.au".

Relevant Principles 224 Section 120(1) of the Trade Marks Act provides that a person infringes a registered trade mark if the person uses as a trade mark, a sign that is substantially identical with, or deceptively similar to, the trade mark in relation to goods or services in respect of which the trade mark is registered. REA did not contend at trial that any of the impugned marks used by Real Estate 1 are substantially identical to its registered trade marks. The only issue pressed under s 120(1) was whether the marks are "deceptively similar". Section 10 of the Trade Marks Act defines a trade mark as deceptively similar to another trade mark "if it so nearly resembles that other trade mark that it is likely to deceive or cause confusion". 225 The legal principles in question are not in dispute and were conveniently summarised by Moore, Sackville and Emmett JJ in Crazy Ron's Communications Pty Ltd v Mobileworld Communications Pty Ltd (2004) 209 ALR 1 at [72]-[81]. The question of deceptive similarity of marks, unlike substantial identity, is not to be judged by a side-by-side comparison. The comparison made is between, on the one hand, the impression based on recollection of the applicant's mark that persons of ordinary intelligence and memory would have, and, on the other hand, the impressions that such persons would get from the respondent's mark. The idea which the mark will naturally suggest to the mind of the person who sees it, will be an important consideration. The effect of spoken description must also be considered so that similarities of both sound and meaning may be important. The usual manner in which ordinary people behave must be the test of what confusion or deception may be expected. Potential buyers of goods are not to be credited with any high perception or habitual caution. On the other hand, exceptional carelessness or stupidity may be disregarded. The course of business and the way in which the particular class of goods are sold, provides the setting and the habits and observations of ordinary people affords the standard. Evidence of actual deception will be of great weight. 226 A mark will be deceptively similar to another mark if there is "a real, tangible danger" of confusion occurring. A mark may be deceptively similar to another mark if its use causes people to wonder whether it might not be the case that the products to which the marks are applied come from the same source. Allowance must be made for the imperfect recollection of a trade mark by people of ordinary memory and intelligence especially those who know one mark but do not know the other: Vivo International Corporation Pty Ltd v Tivo Inc (2012) 294 ALR 661 at [115] (Nicholas J, with whom Dowsett J agreed); and see [77] and [80] (Keane CJ). 227 The threshold for trade mark infringement is lower than that required to establish misleading or deceptive conduct: Puxu at 198 (Gibbs CJ). The wider inquiry of the kind which might be undertaken in a passing off case or where misleading or deceptive conduct is alleged is not appropriate, because the court is not looking at the totality of the conduct: Crazy Ron's at [86]-[87]. The court needs to consider the use of the respondent's mark alone. The fact that a respondent makes it clear by means external to the mark that there is no connection between its product or services and that of the applicant is not relevant: Mark Foy Limited v Davies Coop & Co Ltd (1956) 95 CLR 190 at 205 (Williams J).

Have REA's marks been infringed? 228 It is not disputed that the use of the impugned marks by Real Estate 1 in relation to its websites was use in relation to services in respect of which REA's trade marks are registered, being "advertising of real estate in electronic format" and "provision of online access to real estate information". Nor is it disputed that the marks have been used by Real Estate 1 on its websites and in its advertising and marketing. 229 As to the use by Real Estate 1 of the logos utilised for both its residential and commercial portals, REA submitted that there is a real tangible danger of deception or confusion in circumstances where: (a) the essential or distinguishing feature of each mark is the word contained in the mark, including the ".com.au" domain name elements; (b) the wording for each corresponding mark is identical except for the addition of the numeral "1" and the capitalisation of the word "Estate" or "Commercial"; (c) the marks have been used for identical services; (d) the marks are presented in the same, or a very similar, sans serif font; (e) the marks each use similar devices that serve to illustrate the idea conveyed by the word (in the case of realestate.com.au and realestate1.com.au, a house, and in the case of realcommercial.com.au and realcommercial1.com.au, a factory or collection of commercial buildings); and (f) to the extent the number "1" is a point of difference between REA's and Real Estate 1's marks, it does not appear until the middle of the marks and it is obscured by the way in which it forms part of the device, and it is therefore less likely to be noticed or recalled. 230 On the other hand, Real Estate 1 contended that there are sufficient differences between REA's trade marks and each of Real Estate 1's uses complained of. I will shortly refer to the various points of distinction relied upon. 231 In my view, Real Estate 1's use of the Real Estate 1 logo is not deceptively similar to REA's realestate.com.au trade marks. I have arrived at the opposite conclusion in relation to the realcommercial.com.au trade mark and the use by Real Estate 1 of the realcommercial1 logo. 232 The realestate.com.au logo has arguably three components: a red house in a button device; the words "realestate"; and the term ".com.au". No component on its own is an essential feature as no component is sufficiently prominent. The button device is distinctive but subsidiary. The term ".com.au" is also subsidiary for reasons earlier given. The words "realestate" are descriptive and of themselves do not suggest a connection with a source business or brand. 233 However, the evidence of the widespread consumer recognition of the domain name "realestate.com.au" supports REA's proposition that the essential or distinguishing feature of its logo is the domain name in its entirety. The inclusion of ".com.au" as part of that essential feature is necessary because the highly descriptive nature of "realestate" on its own would not be perceived as sufficient to establish brand identity. It is only the composite expression that is likely to be perceived as providing the brand or trading name of the business being promoted. The impression which the logo creates is that the brand or trading name promoted is "realestate.com.au". The idea that the mark naturally suggests is that a particular domain name (shown in the font and colour that domain names appear in) is also being used as a brand or trading name. 234 In my view, there is no real danger that persons of ordinary memory and intelligence who have seen REA's logo will be deceived or confused when confronted with the Real Estate 1 logo. It is necessary to examine the features of the Real Estate 1 logo, however, ultimately matters of impression rather than side-by-side comparison are determinative. That logo has at least twice and arguably three times the number of components as REA's logo. There are six components as follows: "real"; "Estate"; "1"; the house device; ".com.au"; and the swish. Further the presentation of the logo is plainly differentiated. The font is different including because of the capitalised "E". The colour used and its placement is significantly different. The swish is an added feature. The device utilised is not a button and its colour is dissimilar. The "1" is in bold and tall and stands out. Importantly, the placement of the device in the middle of the logo is a strongly distinguishing feature. That placement has the effect of sidelining ".com.au" and projecting "realEstate1" as the dominant element strongly suggestive of the name of the brand. That effect is reinforced by the swish which underlines "realEstate1" and not ".com.au". 235 The impression left by the logo is that the dominant composite elements or the essential feature is "realEstate1" and that the source business is an internet business that trades by the name "realEstate1". The visibility of the "1" is critical to that impression. Unlike REA's logo, the logo used by Real Estate 1 does not create the strong impression that the brand or trading name of the business is a domain name which includes ".com.au". The different impressions of the brand or trading names projected by the essential features of each logo is significant. Those contrasting features will to some extent be reinforced by the differences in colour and presentation. That all leads me to the conclusion that Real Estate 1's logo does not sufficiently resemble REA's logo to be deceptively similar. I do not consider that, allowing for imperfect recollection, there was a real tangible danger that a person who knew REA's trade mark would have been confused. 236 The main reason I have come to a different conclusion in relation to the realcommercial1 logo is because I consider the essential feature of REA's realcommercial.com.au trade mark to be the words "realcommercial". The impression which persons of ordinary intelligence and memory are likely to take away having seen REA's realcommercial.com.au logo is likely to be based upon the concocted words "realcommercial". The same concocted words are prominent in Real Estate 1's logo where they appear as the strongest element of the logo. Whilst there are differences in font, colour and the use and placement of the house device as well as the existence of additional features, those elements do not do enough, in my view, to avoid the real danger of deception or confusion created by the common essential feature of each logo. For those reasons, I find that the use by Real Estate 1 of its realcommercial1 logo was an infringement of REA's realcommercial.com.au trade mark. 237 It follows from the conclusion just reached that each of Real Estate 1's impugned uses of REA's realcommercial.com.au trade mark must also constitute an infringement. Each of those uses involved the use of "realcommercial" in a manner even less differentiated from REA's trade mark than the use of those words as part of the realcommercial1 logo. Accordingly, I find that each of the impugned uses of REA's realcommercial.com.au trade mark was an infringement of that mark. 238 I do not accept Real Estate 1's contention that for all purposes, an assessment of deceptive similarity in the context of domain names must involve the court ignoring what Perram J in Solahart described at [49] as "the various accoutrement of domain names such as 'www' and '.com.au'". Perram J continued that accoutrements of that kind are not paid substantive attention by consumers. He said: These do not have any impact on the questions either of substantial identicality (from which analysis they must be excised) or on deceptive similarity. 239 Real Estate 1 relied on that observation. However, I do not understand Perram J to have there been making an all encompassing statement of the kind that Real Estate 1's contention suggests. His Honour was there dealing with the use of the trade mark SOLAHART in the domain name "www.solarhut.com.au". The comparison in that case was between a concocted name registered as a trade mark and the use of a very similar word ("solarhut") as part of an internet address. All that I think Perram J was saying was that substantive attention is not paid by consumers to the common accoutrements of an internet address when seeking to identify or distinguish a brand, because of the usually limited function of accoutrements in that context. When used as part of an internet address, "www" and ".com.au" do not ordinarily serve the function of providing cues to brand or product identity. In that context, consumers in pursuit of brand identity are likely to be oblivious to terms that will not be ordinarily recognised as dealing with brand identity. 240 Where, however, a trade mark itself incorporates the accoutrements of a domain name as cues to product or brand identity, the position is different to that considered by Perram J. It is clearly different when the use of the same accoutrements as those incorporated in a trade mark are used not as cues to an address but, like the trade mark itself, as cues to the brand or trading identity of a product or a business. That is what Real Estate 1 did when it used "realEstate1.com.au" as a heading to a sponsored link. An example of such use appeared in the evidence as follows: realEstate1.com.au www.realestate1.com.au/ Our aim is to make searching for real estate pure, simple and fast! 241 In my view, the display of the term "realestate1.com.au" in the heading of a sponsored link would have been regarded by many consumers to be the trading and domain name of the business whose link it was. One of the central distinguishing features of REA's realestate.com.au trade marks is the idea that the term "realestate.com.au" is both a brand name and a domain name at the same time. When Real Estate 1 used "realestate1.com.au" as a trading name, it took up that precise idea. In that context consumers are likely to pay substantive attention to ".com.au" because it serves the function of identifying the brand whose domain name is also being used as a brand. The whole of the domain name is likely to be read or at least scanned. In a circumstance such as that, there was in my view, a real danger of confusion on the part of a consumer familiar with REA's realestate.com.au trade marks. That principally arises because in a scanning process of the kind which can occur on a search results page, the "1", which is not very distinct in the context of a domain name in ordinary type face, is likely to be missed by some consumers. 242 I next need to consider Real Estate 1's use of realestate1.com.au as part of a URL. There is no doubt that a domain name used in a URL can be used to attract consumers to a website from which goods or services are sold and thereby constitute trade mark use: Sports Warehouse at [146]-[153] (Kenny J); Solahart at [50] (Perram J). That however is not the issue here raised. The issue again raised by Real Estate 1 is concerned with the extent to which consumers will take account of the term ".com.au" in the context of its use in a URL. 243 It is to be recalled that REA makes no complaint about the use of "realestate1". Its complaint is necessarily reliant upon the use of that term in combination with ".com.au". The extent to which ".com.au" is likely to be noticed in a URL as an identity cue is therefore a consideration which needs to be examined. In the context now under consideration, Real Estate 1 may again rely upon the observations of both Perram J in Solahart at [49] and Chesterman J in Architects Australia at [21], where both judges thought that the suffix ".com.au" would be given little attention by consumers. Dr Coker considered that in relation to a domain name in a URL, consumers would ignore "www" but may focus on the ".com.au" to gain a geographical cue as to the country location of the website. 244 Ordinarily, the appearance of ".com.au" on a search results page is ubiquitous given the frequent use of the suffix. By reason of its placement in a URL, the use of ".com.au" will ordinarily be regarded as no more than a common geographical cue for the corporation whose address is being described. Unless, in the context in question, a geographical cue is significant, ".com.au" is ordinarily likely to go unrecognised by consumers as a brand identifier. 245 However that observation about the ordinary case does not really address (as neither Perram J nor Chesterman J were called upon to address) a situation where the highly descriptive nature of the second-level domain ("realestate") makes a suffix such as ".com.au" essential to brand or name recognition. Consumers with some familiarity with realestate.com.au as a brand are likely to look beyond "realestate" and to the entire domain name in order to establish identity. A real danger of confusion again arises because in the scanning process which may occur on a search results page, some consumers will miss the indistinctive "1". I have therefore concluded that the use of "realestate1.com.au" as part of an internet address on a search results page, constituted the use by Real Estate 1 of a mark that was deceptively similar to REA's realestate.com.au trade marks. 246 Courts have shown some reluctance to find deceptive similarity by reference to the impact of a common prefix or suffix. In Mond Staffordshire Refinery Co Limited v Harlem (1929) 41 CLR 475 at 477-478, Knox CJ, Gavan Duffy and Starke JJ were influenced by the commonality of the use of "sol" in connection with pharmaceutical preparations to conclude that "Monsol" and "Mulsol" were not deceptively similar. Pinkus J in Miss World (Jersey) Limited v Mrs of the World Pageants Inc [1988] AIPC 90-460 denied interlocutory relief against a "Mrs World" pageant to the organisers of a "Miss World" pageant. His Honour observed that the applicants in that case had chosen for their mark "Miss" and "World", both of which were ordinary and common words. In that situation, Pinkus J considered that: A court would not readily expand the prima facie scope of protection afforded by registration of the relevant mark by conceding a monopoly also over similar titles such as "Mrs" or "Mr"… 247 As my conclusions demonstrate, registration of REA's realestate.com.au marks has effectively given REA a monopoly over two highly descriptive terms when used in combination. Those terms are likely to be the most common terms on a search results page where a search has been conducted for a residential real estate portal. The protection conferred by REA's trade marks over the use of "realestate" and ".com.au" in combination, provides REA with a monopoly over the term "realestate" in circumstances where its rivals seeking also to use "realestate" or a close variant thereof as a second-level domain, do not forego the advantages of using ".com.au" in their domain names. The natural advantage of a domain name which incorporates "realestate" to the commercial success of property portals will be apparent from observations I have already made. There is also a natural advantage in the use of the suffix ".com.au". It is troubling that terms that are highly descriptive of a particular area of commerce and which provide significant commercial advantage should not be readily available for use by all who seek to participate in that commerce. However, in the absence of a successful challenge to the registration of REA's realestate.com.au trade marks, whilst that may be troubling, REA is nevertheless entitled to the protection of the monopoly which has been conferred upon it. 248 For all those reasons, other than in relation to Real Estate 1's use of "realestate1.com.au" as part of the logo for its residential property portal, REA has established each of its claims that Real Estate 1 has infringed its trade marks.

accessorial liability 249 In light of my findings that Real Estate 1 infringed the realestate.com.au trade marks and the realcommercial.com.au trade mark, it is necessary to consider REA's claim that each of the individual respondents were joint tortfeasors in relation to those infringements. 250 Mr and Mrs Luff were directors of both the corporate respondents and the only shareholders of the second respondent. They were also significant shareholders of the first respondent. Both were substantially involved in the day to day running of Real Estate 1's business. Both were involved in the choice of domain names utilised by Real Estate 1. 251 The fifth respondent, Christian Ongarello, is Mrs Luff's adult son. He was employed by the second respondent as a sales representative from November 2009. He was also a director of the first respondent until 18 November 2010. The sixth respondent Bianca Ongarello was a director of the first respondent until 1 June 2011. 252 REA correctly identified the leading Full Court authority on the liability of joint tortfeasors as Keller v LED Technologies Pty Ltd (2010) 185 FCR 449. By reference to that authority, REA contended that the individual respondents used Real Estate 1 as an instrument of their appropriation of REA's goodwill and reputation. 253 That contention relies mainly upon the approach to establishing the liability of a director as a joint tortfeasor which was articulated in the following observations made by Jessup J in Keller at [404]: For a director to be liable because he or she directs or procures his or her company to commit a wrong, the context must be such that the director is effectively standing apart from the company and directing or procuring it as a separate entity. There must be a sense in which the director is using the company as the instrument of his or her own wrong. This is, as I have explained earlier, the original sense in which Atkin LJ's words in Performing Right are to be understood, at least to the extent that they drew upon what Lord Buckmaster had earlier said in Rainham Chemical. Likewise with a joint enterprise, unless the director is conceptually separate from the company, there is, it seems to me, no "jointness" about the enterprise at all, but merely a one-person wrong in which the company is actor. 254 At [405], Jessup J added: in a situation in which the company role and the personal role of the director may be blurred, the formula does highlight the crucial distinction between acts which are done for and in the service of the company and acts which, in addition, are done in the director's own personal capacity - a "non-company capacity" as it were. 255 Similar observations were made by Emmett J in Keller at [83]-[84] as follows: A company cannot act other than through a natural person. In considering whether a natural person is a joint tortfeasor with a company, it is necessary to show something more than that the company acted through that person. Where a person is acting in the capacity of a director, the person will not be liable for the act of the company unless it can be shown that, in so acting, the director was doing something more than acting as a director. The person must do something that makes him or her, in addition to the company, an invader of the victim's rights (see O'Brien v Dawson (1942) 66 CLR 18 at 32-33). The mere fact that a company is small and that the director has control over its affairs is not, of itself, sufficient to make the director a joint tortfeasor with the company (see C Evans & Sons Ltd v Spriteband Ltd [1985] 1 WLR 317 at 329). Infringement by a principal actor, of course, is an objective matter. For a director of a company to be held to be invading the rights of a victim of the company, by reason of the actions committed in the capacity of a director, there must be some mental element involved. Thus, in circumstances where a director can be shown to be making use of a corporation or company as an instrument whereby infringement is perpetrated, such that the director can be seen to be hiding behind the corporate veil, it may be thought that that director is going beyond actions performed merely in the capacity as director. If a company is merely the alter ego of a director, such that there is no real difference between the mind of the officer and the mind of the company, there may well be circumstances where it will be appropriate to conclude that the officer is invading the rights of a victim of the company. 256 In Sporte Leisure Pty Ltd v Paul's International Pty Ltd (2010) 275 ALR 258, Nicholas J referred to Keller and at [118] said: I think the authorities demonstrate that something more will usually be required if a director is to be held liable as a joint tortfeasor than a simple finding that he or she, while acting in that capacity, caused or directed his or her company to perform acts which are later held to be infringing acts. The extent of the director's personal involvement in the commission of the infringing acts is no doubt critical. But his or her state of mind is itself an important aspect of that involvement. That does not mean that knowledge that the relevant acts are infringing is a prerequisite to liability. The authorities are clear in holding that it is not. Even so, a finding that a director who held an honest belief that the acts which he or she directed or procured were not unlawful is a significant consideration telling against the director's liability. 257 In that case, whilst the director had been closely involved in the infringing acts of the corporation, Nicholas J determined that the acts were done in his capacity as a director and at a time when he genuinely believed that the products which were held to have infringed the applicant's trade mark were procured under a licence from the owner of the trade mark. In those circumstances, Nicholas J held that the director was not liable as a joint tortfeasor. 258 I have found that Mr Luff's rebranding exercise which led to the infringing conduct, was done with the intention of pressuring REA to purchase Real Estate 1's business at some later time. I have also been satisfied that Mr Luff was at least in part motivated to adopt realestate1.com.au as Real Estate 1's domain name because he wanted to take advantage of the reputation of realestate.com.au. I could not be satisfied that when Mr Luff organised for "realestate1.com.au" and "realcommercial1.com.au" to be used as brand or trading names by Real Estate 1, he held an honest belief that the use of those names would not infringe the rights of REA. 259 A similar situation was considered by Dodds-Streeton J in Tivo Inc v Vivo International Corporation Pty Ltd [2012] FCA 252. In that case as her Honour outlined at [533]-[534], the sole controller and shareholder of the infringing company was inspired by the applicant's trade mark and intended to obtain an advantage or spin-off benefits through the use of the infringing name. Her Honour was not satisfied that the director held an honest belief that the adoption of the infringing name would not infringe the applicant's right. Despite those findings, Dodds-Streeton J nevertheless found that accessorial liability for any infringement was not established. At [534] her Honour said: Nevertheless, the evidence does not establish that the company was merely the alter ego of Mr Grassia, which he employed as a shelter or personal instrument of his own wrong by standing apart and directing or procuring it to adopt the Vivo name. No dimension of Mr Grassia's role appears materially separate from that of the company. 260 With some reservation, I have come to the view that the evidence in this case requires the same conclusion to be drawn as that reached by Dodds-Streeton J. 261 The evidence does not allow for the conclusion that Mr Luff used Real Estate 1 as a shelter or vehicle for his own wrong. The mere fact that he had control over the affairs of the corporate respondents is not sufficient to establish that he was a joint tortfeasor. The evidence does not establish that he was motivated by personal interests sufficiently separate from those of the interests of the corporate respondents as to establish that the corporate respondents were being used as the instrument of his own wrong doing. 262 The involvement of Mr Luff in the infringing conduct and the extent to which he may be implicated in the infringing conduct, far exceeded that of the other individual respondents. It must follow from my conclusion that REA has not established that Mr Luff is a joint tortfeasor, that accessorial liability is not established in relation to any of the individual respondents.

relief 263 I have found that by using the marks realestate1.com.au and realcommercial1.com.au, Real Estate 1 infringed REA's trade marks in all but one of the different uses alleged by REA. Otherwise, REA's claims expounded at trial have failed. 264 Damage was not the subject of the trial and may now need to be determined in a separate hearing. The form of the other relief sought by REA has not been the subject of submissions and the parties ought to be given the opportunity to make submissions in light of my findings. 265 I will make orders requiring the parties on or before 21 June 2013 to confer and bring in short minutes of orders dealing with: (i) the exchange and filing of submissions as to the extent and form of any relief which ought be granted consequent upon the findings here made; and (ii) the steps to be taken to facilitate any further hearing on the issue of damages. If damages are no longer pursued, the proposed orders as to the filing and exchange of submissions should also deal with submissions as to costs. I certify that the preceding two hundred and sixty-five (265) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Honourable Justice Bromberg.