9.3 The ACT Test, the Yum Model and the first mover advantage
371 Various criticisms can be made of the Yum Model. Matters such as reliance on the New Zealand data, the adoption of the New Zealand benchmark and the labour hours imported into the Yum Model to achieve the 34.5% price lift have been shown by DPL to be validly subject to comment and some criticism. However, these are not the only questions to be asked and it is clear that Mr Sinha and Mr Houston believed that the Yum Model was valid and reliable, as a model.
372 The ACT Test was just that: a test. There is no doubt that the ACT Test can be evaluated in different ways. It had the advantage of involving a number of stores which should have assisted extrapolation to a national average store but those stores were also in a limited and somewhat special geographical area. The use of the ACT as the test location and the use of the ACT stores is not in dispute. However, the parties disagree as to the proper analysis of the results. Yum drew conclusions from the results of all of the stores; DPL concluded that only the Erindale store should have been used to calculate the input of labour hours into the Yum Model.
373 DPL criticises reliance on the ACT Test as it says that the results did not indicate improved profitability but rather a loss and it concludes that Yum actively engaged in a "public relations exercise" by presenting manipulated successful data to the Franchisees, which presented a false picture of profitability. Further, by the time of the decision to implement the VS, Yum knew that the marketing budgeted for the VS would not be forthcoming. Not only did this affect the VS itself which depended on, and included, provision for increased marketing but marketing was also crucial, to the knowledge of Yum, which had contributed a relatively large advertising budget to the ACT Test. However, Yum rejects DPL's theory that expenditure on marketing for the ACT stores must be calculated and then that figure multiplied by the number of stores nationally. Mr Houston gave evidence that his concern was to ensure that Yum could replicate the ACT advertising expenditure actually implemented on a national rollout and explained that "replicate" meant achieve the media outcome achieved in the ACT, not the dollars actually spent. As a result of the circumstances following the implementation of the VS, including the absence of expected extra contributions from Franchisees, the evidence was insufficient to establish whether the marketing spend in the ACT Test and the VS were comparable.
374 It would seem that choices were made by Yum to include and reject data obtained during the ACT Test for the purposes of drawing conclusions as to profitability. DPL has not established that Ms Broad deliberately engineered the results to obtain a false picture of profitability. Ms Broad provided an explanation of decisions that she made and accepted that some data had not been included. While there may be criticisms of her reasons for including some data and not including other data, her decisions, such as which weekly periods to include, have not been shown to be unreasonable or to invalidate her conclusions.
375 Yum also defended its position that the advertising budget provided for the ACT franchisee did not render those results inappropriate to be replicated nationally. While there may be differences of opinion as to the making of predictions, Yum provided an explanation for its view at the time and DPL did not provide evidence to support its theory. DPL has not established that the correct method would have been to take the marketing budget for the ACT Test and just apply it nationally. A key issue was whether the 4% additional LSM should have been included. I am not satisfied that it was inappropriate to exclude it at the time that the calculations were made. Ms Broad did not include the complete 1.8% of LSM and only used 1.5% as a cut-off mark. Ms Broad stated that the reason that she provided for extra marketing was because it was part of overall marketing activity being produced by Ms Syed's team and her understanding was that Yum would not need to replicate all such cost when the marketing strategy was extended to a national level. Had she done so, it could have altered the profit and loss of the ACT franchisee. However, I am not satisfied that this decision was made with some ulterior motive or without reason. There are clearly other factors that would need to be taken into account when assessing this issue, such as those raised by Yum, for example the difference in "reach" of different Franchisees. In the formulation of the VS, there was provision for extra marketing, although the relativities with Domino's were not established. Intervening events, such as Domino's prior market entry and the failure to get Adco approval and Franchisee marketing contributions make it impossible to conclude that the failure was due to the VS itself.
376 There is no dispute that it was appropriate for Yum to conduct a test of the VS. There is no good reason to conclude that testing in the ACT Test was inappropriate, or that the test was poorly conducted. There will always be issues with such a test and it will not always accurately predict whether it will be successful on a national basis or for other franchises.
377 DPL relies on Mr Potter's analysis. Mr Potter's analysis was based on different parameters and, in particular, on the results of the Erindale store. Mr Gower challenged Mr Potter's analysis but did not proffer an alternative or provide a rigorous analysis. Mr Potter's conclusion was that the additional labour required to obtain the necessary 34.5% uplift was four times that stipulated by Yum; even if this figure, of 52 hours, is too high, alternative calculations based on Mr Potter's assumptions still result in significantly higher additional labour hours than the provision in the Yum Model. DPL maintains that the Erindale store should have been used as the basis for the Yum Model and for the labour hours required to provide the uplift. Yum disagrees and says that the Erindale store is no more representative of the national average store than are the other stores in the ACT with respect to labour usage and efficiency. DPL says that this statement is a non-sequitur. DPL says that what made Erindale representative was its sales turnover and it had a relatively close approximation to the key labour statistics of the national average, certainly closer than the other ACT stores. Yum says that the fact that it was close to the national average says nothing about whether it was representative of the national average store with respect to labour usage or efficiency.
378 I am not satisfied that DPL has established that only the Erindale store should have formed the basis of a model of the national average store or that Yum was in error in using all of the stores the subject of the ACT Test for the purposes of the Yum Model. It did not lack reason to utilise the ACT Test to examine the strategy over a range of stores in a geographical area, so as better to have an average of stores for the characterisation of a national average store for a model of the effect of a strategy.
379 DPL does not agree that the 13 hours provided for in the Yum Model was a reasonable estimation of variable labour hours which would be required to provide for a 34.5% uplift in transactions. That estimate was made by Mr Sinha. He gave a detailed explanation of his reasoning. That explanation was not shown to be unreasonable, nor did DPL show that the decision was made in bad faith or recklessly. DPL's case rests upon an allocation for labour hours greatly in excess of Mr Sinha's estimate, which was based on his reasoning in section 5.4.4 above. DPL has not demonstrated an error of the order of magnitude it advances. Mr Sinha's estimate was based on Mr Sinha's own experience, including as to the time taken to make a pizza, the use of labour, labour availability in a store across the week and information available to him from New Zealand and the ACT Test. It may have been lacking in detailed analytics, including interrogating the data to separate driver hours, as carried out by Mr Potter, but Mr Sinha is not a qualified accountant; he worked his way up in the Pizza Hut business. He explained his own experience as to the time taken to make a pizza and as to the information that he received from the ACT franchisee as to the use of delivery drivers.
380 Mr Sinha was adamant that the provision in the Yum Model of 13 additional hours was reasonable. Mr Sinha gave evidence as to his reliance on the New Zealand data and what he drew from the ACT Test but he also drew on his own experience, including as a pizza maker in a Pizza Hut franchise. Mr Sinha explained that he did not in fact rely on the New Zealand data alone for determining the labour hours for the Yum Model. Even accepting that he used the incorrect data point from New Zealand by way of reference, he did not simply insert that into the model but also relied on the ACT Test results to establish the correct data point. Whether or not 5.6 was the appropriate New Zealand benchmark for measuring labour efficiency during the ACT Test, New Zealand merely represented an imperfect comparator. For example, RBNZ was the master franchisee in New Zealand and Pizza Hut has a greater market share in New Zealand than does Domino's. The fact is that Mr Sinha says that he used it merely as a comparator and relied on the ACT Test results and his own experience and that evidence was not shaken.
381 DPL asserts, in effect, that the figure for labour hours was a number randomly allocated to be used as a balancing item in Yum's calculations. The differences between Yum's labour hours calculations, relied upon by DPL, are those that appeared in iterations of the Yum Model, including 9 and 13 hours. Yum points out that this represents a difference of $60 in labour costs. Mr Sinha gave an explanation for his choice of allocation of labour hours, although, there was no real explanation as to the variation in the final presentation of the Yum Model. However, Yum's submission is that DPL did not establish that 9 hours was the more appropriate figure. In the light of Mr Sinha's evidence, DPL has not established that proposition.
382 DPL points to different versions of the Yum Model that preceded the model presented to Mr Houston and on which he based his decision. Some of the versions in evidence were the work of Mr Purcell, who was not called to give evidence. Those versions contained different figures for labour hours and the required or estimated transaction growth percentage. DPL relies on these versions to submit that the Yum Model was not created rigorously or bona fides to represent what was achievable. Yum maintains that these earlier versions represented the structure of the Yum Model at an early stage, rather than a real version of the Yum Model to be presented to Mr Houston and the Franchisees.
383 Minds may differ as to whether Mr Sinha with practical experience or an accountant with theoretical qualifications would be better placed to determine such parameters for the Yum Model based on the information available. That may, in turn, also be affected by the use of the Yum Model. Mr Sinha and Mr Potter came to different conclusions, in part based on different use of the ACT Test data and on the breadth of those data compared to the data for one store. It is not uncommon for different people with different perspectives to have different opinions as to the parameters, efficacy and applicability of a model. Even accepting that Erindale represented what should happen with a well-run store, that does not mean that only Erindale data should apply to a model of the national average store and other data rejected. It was not shown to be unreasonable or negligent to use Mr Sinha rather than an accountant. DPL has not shown that Mr Sinha's reasoning was not open or that his determination was unreasonable or made in bad faith or recklessly. Rather, Mr Sinha was satisfied that his calculation of 13 additional labour hours was reasonable and provided an appropriate input into the Yum Model.
384 The fact that criticisms can be made, for example that Mr Sinha used the New Zealand benchmark rather than the underlying New Zealand data, may mean that he should have analysed those data more carefully but it does not mean that the benchmark represented an unreasonable figure. Ultimately, Mr Sinha formed the view that it was consistent with his experience and with the ACT data.
385 Similarly, DPL recognises that a business can be modelled in different ways. It prefers Mr Potter's analysis and use of a product/cost model. Yum does not accept the validity of Mr Potter's analysis nor of the assumptions that he has made. Yum prefers Mr Sinha's and Mr Gower's and a break even model. Each party challenges the assumptions made in the opposing model. However, DPL has not shown that the Yum Model was developed unreasonably or in bad faith or negligently.
386 Mr Potter has shown that additional labour hours can reasonably be calculated from the data to be significantly higher than 13 hours, especially if it can be accepted that, as demonstrated in the ACT Test and as explained by Mr Sinha, different stores adopt different usage and record of payment of drivers, some of whom also work in the store and are then paid on an hourly basis, whereas other stores use drivers that only do a guaranteed number of deliveries. However, this demonstrates that different models could reasonably have been created, especially where decisions are made by former pizza store managers and compared to those of analytical accountants. It does not necessarily follow that Mr Sinha's Yum Model was flawed or, if so, he should have appreciated that fact.
387 Mr Potter calculated that a $4.95 Classics pizza price point was unprofitable but failed to take into account other aspects of the VS, including other pizza range price points and subsequent changes to the prices following the implementation of the VS. His calculations are challenged by Yum and Yum submits, in effect, that in any event Mr Potter's conclusions are not connected to the Yum Model, the design of which does not correlate with Mr Potter's methodology. In any event, Yum's evidence is to the effect that Mr Houston and ultimately Ms Broad, as well as Mr Smith, were of the view that the VS as a whole, including the $4.95 Classics pizza and the uplift in sales, would increase Franchisee profits.
388 DPL contends that the Yum executives knew or ought to have appreciated that the VS as a whole would be unprofitable or loss-making for the Franchisees. Nevertheless, DPL's submission that the implicit bargain is that Yum will not impose 'an unprofitable price' that will negatively impact on the Franchisees' profitability is somewhat simplistic. DPL emphasises the price of a Classics pizza, being the pizza for which the $4.95 maximum price was imposed. However, the point of the VS was to bring about a 34.5% sales uplift, not just by the sale of more Classics pizzas but also by the increased sale of other pizzas in the range and side orders, together with increased deliveries for which there was a delivery charge. DPL's focus on the $4.95 price point of the Classics range has not taken into account the variation in the mix of pizzas before and after the implementation of the VS. Before the implementation of the VS, the Classics range was one of four ranges of pizza in the mix; following the implementation, the mix was reduced to two ranges. There were also sides and delivery fees to be considered as part of the total sales. The $4.95 Classics pizza could be viewed as a "loss leader" to bring about a substantial increase in overall sales and hence increased profitability for both Yum and the Franchisees.
389 DPL's submissions that the ACT data were manipulated by Yum in order to present a false picture of profitability of the ACT Test to Franchisees, while concealing the data from Jagot J seem to be linked with its contention that Yum wanted to implement the VS for some purpose other than assisting the Australian Pizza Hut businesses. DPL offers a reason for this: that Yum was acting under direction from Yum US, who wanted the VS implemented in Australia for its own reasons, not concerned with Australian profitability.
390 DPL has not established such a conspiracy. The evidence, including that of Mr Houston and Ms Broad and the documentary evidence as a whole, demonstrates that the decision was that of Mr Houston. Certainly he had discussions with executives of Yum US and took account of their views and sought alignment with the FPC to ensure that Yum received the extra $1 million it sought in funding but he did not act under their direction. He acted in what he believed was the best interests of the Australian enterprise - rightly or wrongly. It does not make sense, as DPL's theory would encompass, that Mr Houston would make a decision deliberately to diminish the profitability of the Franchises on which the Yum business depends. It is true that in the short term, payments by Franchisees to Yum would increase with the predicted uplift in transactions but it still does not make commercial sense to make a decision for that purpose where it would be only short term and would result in increased bad debts from Franchisees and franchises failing. That would not assist Yum commercially, as the value of a franchise would decrease.
391 I reject DPL's contention that the decision to devise and implement the VS was made by Yum US and not by Yum in Australia. The evidence, both written and oral, is consistent with, and supports the conclusion that, the decision being made here. I accept Yum's detailed submissions as to the characterisation of the relevant events, discussions and emails. Certainly, there were discussions with Yum US executives and the idea of a VS including velocity pricing may well have come from there but the decision, ultimately, was clearly made by Mr Houston with his executives in Australia and not by a decision, direction or approval from Yum US.
392 It was clear from the Help! email that the decision was Mr Houston's. He used Mr Bergren in the US as a sounding board but the decision was his. The idea to consider the VS in Australia may have come from Yum US and its executives participated in discussions about it and Yum US may have provided funding for it but the decision to implement was made by Yum. DPL has not established the contrary.
393 DPL says that a failed franchise is in Yum's interests because it is able to recover debts from the existing Franchisee and then sell the franchise at a profit to Yum and recoup more by way of initial fees payable under the IFA. That speculation is not borne out by such evidence as is available. By the time that Yum writes off the bad debt and refurbishes or otherwise prepares a store for resale, the evidence is that Yum does not make a profit and may, indeed, make a loss.
394 As to the Yum Model, the first matter to note is that it was a model, necessarily containing assumptions and including parameters that were chosen but could be varied to arrive at different outcomes. It was a tool. If one input is changed, it necessarily changes the inputs in other cells of the model. It is not uncommon, and was a feature here, that in arriving at the final version of the model, inputs are changed in an iterative process to see what effect they have and to determine the appropriate parameters. In the end, however, it remains a model, a theory to assist in determining an outcome based on certain parameters. Inherently, it contains a measure of hypothesis and speculation. It will not necessarily be universally applicable. In this case, the Yum Model was for a "National Average" outlet and clearly this means that the parameters will not be applicable to all outlets.
395 Whether or not there could be other models that could have been prepared, the Yum Model was not a pricing model. Nor was it a model to predict the outcome of the implementation of the VS. It was a "break even" model, to ascertain the level of increase in transactions required for the national average store to retain the same EBITDA level of profitability after the introduction of the VS. This being the case, Yum was considering the profitability of the Franchisees and the bargain embodied in the IFA.
396 I reject the submission that Mr Houston's decision to implement the VS was made in bad faith, or on the orders of Yum US, or without consideration of the Franchisees, or simply to increase Yum's share of increased turnover irrespective of whether it was accompanied by increased profit or loss by the Franchisees. As viewed today, Mr Houston may not have been totally adequate for his role in steering Yum through the task of deciding on the best and most accurate Yum Model and whether or not to implement the VS. However, he was the CEO of the company, entrusted with oversight of those matters and he sought to fulfil his tasks to the best of his ability. In making the decisions, he was entitled to delegate appropriate matters, such as the creation of the Yum Model to other Yum employees. For example, Mr Sinha was, to Mr Houston's knowledge, sufficiently experienced to help to create the model. The fact that Mr Sinha had no accounting qualifications did not seem to affect Mr Houston's faith that, as an experienced Pizza Hut employee who had risen to National Operations Manager of Yum, he could rely on Mr Sinha to provide appropriate input into the Yum Model.
397 It would only be speculation to consider what would have happened had the Franchisees not applied for an injunction, or to consider what would have happened had Domino's not entered the market. Each event occurred and had consequences. I accept Yum's evidence that it also perceived that the first mover advantage was a relevant and important factor in the launch of the VS.
398 Decisions made subsequent to the launch of the VS were clearly attempts to recover from the losses being experienced by the Franchisees, while at the same time offering comparable prices and ranges to those offered by Domino's.
399 The evidence is that Domino's did respond to the new pizza prices in the ACT by advertising in the ACT, including on television. Mr Creedy warned Mr Houston not to rely on the lack of response in New Zealand and that Domino's would respond in Australia. I do not accept Yum's evidence that it was unaware of Domino's response in the ACT, although that matter may not have been reported to the Yum leadership team or Mr Houston. It is not credible that a responsible marketing manager who was aware of the importance of the major competitor's response would fail to monitor the media broadly upon the implementation of the ACT Test. In any event, when Mediacom advised Ms Syed of the Domino's advertisement on 13 June 2014, she did advise the Yum leadership team. This was in advance of the launch of the VS and by then Yum knew that Domino's was likely to respond immediately to the VS and that it did respond in the ACT. However, this does not mean that Yum failed to believe in the first mover advantage, although Yum appreciated, or should have appreciated, that the first mover advantage would be short-lived or diminished.
400 Hindsight grants the viewer 20/20 vision of the success or failure of a proposed strategy. Further, it enables analysis of the information fed into the Yum Model, of the success or failure of the ACT Test and how it was evaluated, and of the fact that the results were not as predicted. It also takes account of the fact, which was not part of the Yum Model, that Domino's would be first to market with an all day, every day $4.95 pizza. That is, the VS that was in fact implemented as part of a marketing strategy was not the same as the VS that was devised.
401 Once Domino's launched, Mr Houston had to decide whether Pizza Hut could afford not to implement the available strategy which was ready for implementation and for which media had been arranged, or whether the best business decision was to implement it and to try to avoid Domino's being the only one with this offer on the market.