FEDERAL COURT OF AUSTRALIA

REA Group Limited v Fairfax Media Limited [2017] FCA 91

REASONS FOR JUDGMENT

A.    INTRODUCTION

1    In February 2016 the respondent, Fairfax Media Limited (Fairfax), published advertisements (the Advertisements) which promoted a mobile telephone application (app) operated by its wholly owned subsidiary Domain Group (Domain). The applicant, REA Group Limited (REA), alleges that the publication of the Advertisements constitutes misleading or deceptive conduct or conduct which was likely to mislead or deceive and that Domain made false or misleading representations in connection with the supply of services in breach of ss 18 and 29 of the Australian Consumer Law (ACL) in Schedule 2 of the Competition and Consumer Act 2010 (Cth) (CCA). For convenience I will usually refer to the respondent as “Domain”.

2    Domain is a digital and print media business which operates the property listings website www.domain.com.au, the associated mobile website and the Domain mobile telephone application (the Domain app). REA is a publicly listed digital media business which, amongst other things, owns and operates the property listings website www.realestate.com.au, the associated mobile website and the Realestate.com.au app (the REA app). The two companies are trade rivals in running what are, essentially, classified advertising businesses. Each of them advertises properties for sale or lease (property listings) on its app and website, with the same property listings displayed on the app as on the website. Their apps and websites are free to use and their revenue is principally derived from fees charged to real estate agents for the property listings. They directly compete for this revenue and they are the two dominant players in the market.

3    Domain commenced publishing the impugned Advertisements in February 2016 in a campaign with two distinct phases in New South Wales (NSW), Victoria and Queensland. In relation to the first phase of the campaign REA’s complaint relates to claims that Domain has the “#1 property app in Australia”, “the most property listings in Sydney” and “the best property listings in Melbourne”. In relation to the second phase of the campaign REA’s complaint relates to claims that the Domain app is “Australia’s highest rated property app”.

4    REA seeks relief in relation to six Advertisements which I will call the Second, Third, Fourth, Fifth, Sixth, and Seventh Advertisements. REA abandoned its claim in relation to the First Advertisement.

5    In broad summary, the Second and Fifth Advertisements made the claim “the most property listings in Sydney”. I consider they conveyed the representation that “the Domain app and website has more property listings in Sydney than any other comparable app or website”. The Second Advertisement also conveyed the related representation that the Domain app is the “#1 property app in Australia” because it allows the user to view “the most property listings in Sydney”. The evidence shows that the Domain app and website did not have more property listings overall in Sydney than the REA app and website at the relevant time. I consider that in publishing the Second and Fifth Advertisements Domain made false or misleading representations and it engaged in misleading or deceptive conduct or conduct which was likely to mislead or deceive.

6    However, I do not consider that in publishing the balance of the claims in the Advertisements Domain made false or misleading representations, nor that it engaged in misleading or deceptive conduct or conduct which was likely to mislead or deceive. Broadly this is because:

(a)    the Third Advertisement made the claim that “the best property listings in Melbourne are on Domain”. It also conveyed the related claim that the Domain app is the “#1 property app in Australia” because it allows the user to view “the best property listings in Melbourne”. In my view these indefinite claims of superiority would be understood by the ordinary or reasonable member of the target audience as just advertising puffery. They would not convey a representation about Domain’s app upon which he or she would rely;

(b)    the Second and Third Advertisements made the claim that the Domain app is the “#1 property app in Australia”. In my view this claim, considered as a distinct claim, would also be understood by the ordinary or reasonable reader as just advertising puffery which had little content and did not convey a definite representation as to the superiority of the Domain app. Nor do I consider this claim conveyed the representations REA alleged, namely that “the Domain app is the most widely used or commonly used property app in Australia” or that “the Domain app has been ranked as the #1 property app in Australia by an independent, objective and reliable source”;

(c)    the Fourth, Sixth and Seventh Advertisements made the claim that the Domain app is “Australia’s highest rated property app”. It conveyed the representation that the Domain app has been rated more highly than any other Australian property app, as well as the representation that it was so rated by consumers. Contrary to REA’s contentions, in my view those representations were not limited to just the most recent update of the app. On that basis the representations had a legitimate basis. The evidence shows that at the relevant time the Domain app was the highest rated property app of its kind by reference to consumer ratings of all versions of the app on the Apple App Store and the Google Play Store (the All Versions Rating). I do not consider this claim conveyed the representation that it had been so rated by an independent, objective and reliable source.

7    I thank the parties for the high quality of their oral and written submissions, upon which I have directly drawn at various points.

B.    THE EVIDENCE

8    REA relies on the following material:

(a)    affidavits of Andrew Michael Rechtman, Executive General Manager - Residential of REA Group Ltd, affirmed 12 February 2016, 15 February 2016, and 17 February 2016;

(b)    an affidavit of Henry Joshua Ruiz, Chief Digital Officer of REA Group Ltd, sworn 1 March 2016. Mr Ruiz was cross-examined;

(c)    an affidavit of Jennifer Clare Blake, Digital Analytics Manager of REA Group Ltd, affirmed 1 March 2016;

(d)    an affidavit of Cheron Deysolong Delmoro, Account Manager of REA Group Ltd, affirmed 1 March 2016;

(e)    affidavits of Simon William Hope, Senior Manager – Data Services of REA Group Ltd, affirmed 1 March 2016 and 7 March 2016. Mr Hope was cross-examined;

(f)    an affidavit of Mark James McCowan, Partner in the firm Corrs Chambers Westgarth, affirmed 2 March 2016;

(g)    an affidavit of Alistair John Newton, Senior Associate in the firm Corrs Chambers Westgarth, affirmed 7 March 2016;

(h)    affidavits of Gabriella Anne Bornstein, Solicitor in the firm Corrs Chambers Westgarth, affirmed 7 March 2016 and 17 March 2016;

(i)    an affidavit of Claire Lynnette Morris, Solicitor in the firm Corrs Chambers Westgarth, affirmed 17 March 2016;

(j)    an affidavit of Benjamin Phillip Morling, Social Media Coordinator of REA Group Ltd, affirmed 17 March 2016;

(k)    an affidavit of Shaun Michael Gilbert, Account Manager of REA Group Ltd, affirmed 17 March 2016; and

(l)    the report of the Court appointed expert, Owain Stone of KordaMentha Forensic, dated 21 March 2016.

9    Domain relies on the following material:

(a)    affidavits of Melina Cruickshank, Chief Editorial and Marketing Officer of Domain Group, affirmed 16 February 2016, two affidavits on 1 March 2016, 21 March 2016, and 31 March 2016. Ms Cruickshank was cross-examined;

(b)    an affidavit of Sarah Myers, Head of Brand and Communications of Domain Group, affirmed 1 March 2016;

(c)    an affidavit of Henrique Marassi, Senior Product Manager of Domain Group, affirmed 1 March 2016;

(d)    affidavits of Clinton McNabb, Audience Insights Manager of Domain Group, affirmed 1 March 2016 and 8 March 2016;

(e)    an affidavit of Gail Hambly, Group General Counsel and Company Secretary of Fairfax Media Limited, sworn 1 March 2016; and

(f)    the report of Mr Stone.

C.    THE LEGISLATION AND LEGAL PRINCIPLES

10    Section 18(1) of the ACL provides:

Misleading or deceptive conduct

A person must not, in trade or commerce, engage in conduct that is misleading or deceptive or is likely to mislead or deceive.

11    Section 29(1) of the ACL relevantly provides:

False or misleading representations about goods or services

A person must not, in trade or commerce, in connection with the supply or possible supply of goods or services or in connection with the promotion by any means of the supply or use of good or services:

(a)    make a false or misleading representation that goods are of a particular standard, quality, value, grade, composition, style or model or have had a particular history or particular previous use; or

(b)    make a false or misleading representation that services are of a particular standard, quality, value or grade;

…

12    There is no real contest between the parties as to the principles to be applied, and their differences boil down to the application of the principles to the facts of the present case. I set out the relevant principles below.

13    There is no meaningful difference between the terms “false or misleading”, “mislead”, “misleading or deceptive” and “mislead or deceive” as they appear in ss 18 and 29: Australian Competition and Consumer Commission v Coles Supermarkets Australia Pty Ltd (2014) 317 ALR 73; [2014] FCA 634 (Coles) at [40] (Allsop CJ); Australian Competition and Consumer Commission v Dukemaster Pty Ltd [2009] FCA 682 at [14] (Gordon J). I will treat the terms interchangeably.

14    The Court must determine two main issues:

(a)    Whether the pleaded representations were conveyed by the relevant conduct; and

(b)    If yes, whether the representations were misleading or deceptive or were likely to mislead or deceive, or were false or misleading. This is a question of fact.

Campomar Sociedad Limitada v Nike International Ltd (2000) 202 CLR 45; [2000] HCA 12 (Campomar) at [100]-[101] (Gleeson CJ, Gaudron, McHugh, Gummow, Kirby, Hayne and Callinan JJ); Google Inc v Australian Competition and Consumer Commission (2013) 249 CLR 435; [2013] HCA 1 at [89] (Hayne J).

15    Because the Advertisements were published to the public at large rather than being directed at specific individuals the Court must assess the conduct by reference to the class or classes of persons to whom the conduct was directed. The following steps are appropriate: see .au Domain Administration Ltd v Domain Names Australia Pty Ltd (2004) 207 ALR 521; [2004] FCA 424 (.au Domain) at [12]-[26] (Finkelstein J); Australian Competition and Consumer Commission v Telstra Corporation Limited (2007) 244 ALR 470; [2007] FCA 1904 at [14]-[17] and [19] (Gordon J); Australian Competition and Consumer Commission v SMS Global Pty Ltd [2011] ATPR 42-364; [2011] FCA 855 (Murphy J).

16    First, it is necessary to identify the relevant section or sections of the public at which the Advertisements were directed: Parkdale Custom Built Furniture Pty Ltd v Puxu Pty Ltd (1982) 149 CLR 191 at 199; [1982] HCA 44 (Puxu) (Gibbs CJ); Finucane v New South Wales Egg Corporation (1988) 80 ALR 486 at 516 (Lockhart J).

17    Second, consider who comes within that target audience. Advertisements published to the public at large will include a broad range of people including the astute and the gullible, the intelligent and the not so intelligent, the well-educated and the poorly educated: Puxu at 199 (Gibbs CJ).

18    Third, in cases where representations are made to the public, it is necessary to enquire by reference to a hypothetical ordinary or reasonable member of the target audience whether that hypothetical person is likely to be misled or deceived, excluding reactions to the representation that are extreme or fanciful: Campomar at [103]-[105].

19    There is some debate about whether it is necessary for the applicant to show that a significant proportion of the relevant class of persons were misled or were likely to be misled by the relevant conduct. In Telstra Corporation Ltd v Phone Directories Company Pty Ltd (2014) 316 ALR 590; [2014] FCA 568 at [397]-[399] I said:

In .au Domain at [21]-[26] Finkelstein J criticised the authorities which set out a requirement that for a contravention of s 52 to be established it must also be shown that a significant number of the members of the target class were misled or deceived or were likely to have been so. In his Honour’s view that requirement was wrongly imported from the law of passing off, is not relevant in relation to s 52, and is inconsistent with Campomar.

I am attracted by Finkelstein J’s approach in .au Domain but, since that case several Full Courts have maintained the approach that it is necessary to establish that a “not insignificant” proportion of the class are likely to have been misled or deceived: see National Exchange Pty Ltd v Australian Securities and Investments Commission (2004) 61 IPR 420 per Jacobson and Bennett JJ at [70]-[71], per Dowsett J at [23]; Hansen Beverage Co v Bickfords (Aust) Pty Ltd (2008) 171 FCR 579 at [46] per Tamberlin J and at [66] per Siopis J; Bodum at [206]-[210] per Greenwood J with Tracey J approving.

In most cases whichever of the tests is applied the result is likely to be the same. An assessment that the hypothetical ordinary and reasonable member of the target class is likely to have been misled or deceived will usually exclude the insignificant number of class members whose reaction is extreme or fanciful. I agree with Dowsett J in National Exchange at [23] in which he described the “significant proportion” test as simply an alternative way of expressing the test in Campomar. But, out of an abundance of caution, having assessed the likely effect of the conduct on the hypothetical ordinary or reasonable member of the target class, excluding reactions to it that are “extreme” or “fanciful, I will also assess whether a not “insignificant number” of that class are likely to have been misled or deceived.

I take the same approach in the present case.

20    Fourth, whether the hypothetical member of the target audience is likely to be misled or deceived may involve questions as to the knowledge properly to be attributed to the members of the target audience: Australian Competition and Consumer Commission v TPG Internet Pty Ltd (2013) 250 CLR 640; [2013] HCA 54 (TPG) at [45], [53] (French CJ, Crennan, Bell and Keane JJ).

21    Fifth, it is unnecessary to prove that a misconception has arisen because of the alleged misleading or deceptive conduct. Evidence that someone was actually misled or deceived by an advertisement is not required: Taco Co of Australia Inc v Taco Bell Pty Ltd (1982) 42 ALR 177; [1982] FCA 170 at 199 (Deane and Fitzgerald JJ). In circumstances where the advertisements were directed to the public at large the absence of such evidence is not of great significance: Coles at [45]. On the other hand, evidence that a person has been misled by an advertisement is not conclusive that its publication constitutes misleading or deceptive conduct.

22    The representations must be considered as a whole and in the context of all the surrounding circumstances. It is wrong to select some words only and ignore others which provide the context which gave meaning to the particular words, and the same is true of acts: Puxu at 199; Coles at [41].

23    If one or more of the reasonably available different meanings conveyed to the target audience was misleading, the conduct may have been misleading, deceptive or false and misleading: Tobacco Institute of Australia Ltd v Australian Federation of Consumer Organisations Inc (1992) 38 FCR 1 at 50; [1992] FCA 962 (Tobacco Institute) (Hill J). In Tobacco Institute Hill J said:

Where, as in the present case, the advertisement is capable of more than one meaning, the question of whether the conduct of placing the advertisement in a newspaper is misleading or deceptive conduct must be tested against each meaning which is reasonably open. This is perhaps but another way of saying that the advertisement will be misleading or likely to mislead or deceive if any reasonable interpretation of it would lead a member of the class, who can be expected to read it, into error: Keehn v Medical Benefits Fund of Australia Ltd (1977) 14 ALR 77 at 81 per Northrop J and cf the approach taken by Mason J in Parkdale.

24    In Coles at [46]-[47] Allsop CJ approved Hill J’s remarks and said:

Where advertising material uses simple phrases and words evoking attractive notions, but without necessarily precise meanings, ambiguity or reasonably available different meanings may well arise. Context and the “dominant message” will be important. If one or more of the reasonably available different meanings is misleading, the conduct may well be misleading or deceptive, or false and misleading.

25    It is appropriate to examine comparative advertising more critically than other advertising because the ordinary or reasonable reader is “less likely than otherwise to regard what is said or written as mere exaggeration, stated with an excess of enthusiasm”: Hospitals Contribution Fund of Australia Ltd v Switzerland Australia Health Fund Pty Ltd (1987) 78 ALR 483 at 484; [1987] FCA 395 (HCF) (Fox, Davies and Morling JJ). Where the advertising is comparative, the comparison may be misleading if material that would otherwise make the comparison fair is omitted. Although the public are accustomed to the puffing of products in advertising, comparative advertising requires that the advertiser take particular care not to mislead. An unfair comparison may be misleading: Duracell Australia Pty Ltd v Union Carbide Australia Ltd (1988) ATPR 40-918 at 49,861 (Burchett J); Hoover (Australia) Pty Ltd v Email Ltd (1991) 104 ALR 369 at 375; [1991] FCA 646 (Gummow J).

26    Where the representations are conveyed by an advertisement the dominant message of the advertisement may be crucial to assessing whether it was misleading or deceptive. The Court must be mindful that many readers will read the advertisement fleetingly, will not closely study its constituent parts, and will only absorb its general thrust. The impression or thrust conveyed to a viewer, particularly the first impression, its critical in determining the representation conveyed rather than an analysis of the (sometimes) cleverly crafted constituent parts: Tobacco Institute at 4 (Sheppard J); Australian Competition and Consumer Commission v TPG Internet Pty Ltd [2011] FCA 1254 at [38], [43] (ACCC v TPG) (Murphy J); TPG at [45].

27    As I said in ACCC v TPG at [58]-[59], where an advertisement has a dominant message:

Any purported corrective or qualifying information must be sufficiently clear and sufficiently prominent if it is to prevent an inaccurate primary message from being misleading or likely to mislead: Medical Benefits Fund of Australia Ltd v Cassidy (2003) 135 FCR 1 at [37]; National Exchange Pty Ltd v Australian Securities and Investments Commission [2004] FCAFC 90 at [51]; Singtel Optus Pty Ltd v Telstra [2004] FCA 859 at [41]; Australian Competition and Consumer Commission v Global One Mobile Entertainment Pty Ltd [2011] FCA 393 at [50].

The degree of prominence required to dispel a false dominant message increases with its potential to mislead: ACCC v Signature Security Group [2003] FCA 3 at [27]; National Exchange at [55]; ACCC v Boost Tel Pty Ltd [2010] FCA 701 at [77] to [81]. Where the disparity between the primary message and the true position is great, attention must be drawn to the true position in the clearest possible way: National Exchange at [55]. In Boost Tel at [77] to [81] the Court made a distinction between information which operates to fairly qualify a primary message, and information which undermines the substance and integrity of the primary message. The latter requires greater prominence if it is to be corrective. The question is one of degree and impression. First impressions are important: Medical Benefits Fund at [41].

28    The Court can undertake an assessment of an advertisement based on its ordinary words in context without recourse to witness testimony. It is ultimately a question of fact to be decided by the Court in the particular circumstances of the case and it should not be complicated or over intellectualised: Australian Competition and Consumer Commission v Telstra Corporation Ltd (2004) 208 ALR 459; [2004] FCA 987 (ACCC v Telstra) at [49] (Gyles J).

29    The law is not intended to protect people who fail to take reasonable care to protect their own interests. The ordinary or reasonable member of the public is expected to possess a degree of robustness in relation to claims made in advertising and is accustomed to some puffing of the products and services advertised: Campomar at [102]; Stuart Alexander & Co (Interstate) Pty Ltd v Blenders Pty Ltd (1981) 37 ALR 161 at 164-165; [1981] FCA 169 (Stuart Alexander) (Lockhart J); Singtel Optus Pty Ltd v Telstra Corp Ltd [2004] FCA 859 at [66] (Jacobson J); ACCC v TPG at [39]. Nevertheless, an advertisement that fails to deceive a more wary reader may still be misleading or deceptive.

30    Conduct is misleading or deceptive or likely to mislead or deceive if it has the tendency to lead another person into error. There must be a sufficient causal link between the conduct and the error by the persons to whom the conduct is directed: TPG at [39]. Causing consumers to be confused or to ask questions is insufficient.

D.    THE TARGET AUDIENCE

Identifying the target audience

31    The Advertisements were directed at a broad cross-section of the public rather than to specific individuals, and they were primarily aimed at motivating the reader to download and use the Domain app.

32    Domain’s app is free to download and use and its website is free to use (as are REA’s app and website). Property vendors or lessors and real estate agents choose to list properties on such apps and websites because of the exposure that the listings receive. The evidence shows that the extent to which consumers engage with a property app or website by using it to search for, enquire about, and ultimately buy or lease properties are key measures of that platform’s success. The more properties that are listed on a particular app or website, the more likely that potential property purchasers or lessees will use that service to search for properties. The evidence shows that real estate agents and prospective property vendors or lessors will make their advertising decisions in large part based on which service is likely to reach the greatest audience. Simply put, the greater the number of persons who use the app or website, the greater the number of real estate agents who are likely to advertise property listings on that app or website. Thus there is a circular relationship between the number of property listings and the usage of the platform, with each influencing the other. The evidence indicates that Domain has recently been engaged in a campaign to significantly increase usage of its app, and it has enjoyed some success in doing so.

33    The Second, Third and Fourth Advertisements were published in The Sydney Morning Herald, The Age and The Sunday Age respectively, which have a large readership in NSW and Victoria. The Fifth and Sixth Advertisements were published on billboards at various busy locations in NSW, Victoria or Queensland. The Seventh Advertisement was published on Facebook. I am satisfied that the Advertisements were published to a great number of people in NSW, Victoria and Queensland, but the evidence does not allow an estimate of the number.

34    The Advertisements were not directed at persons with little or no interest in the property market and those persons fall outside the target audience. They would have given the Advertisements little more than a passing glance and they are unlikely to have absorbed any message. They will not have been misled by the Advertisements.

35    Domain contends that there are two relevant classes for the purposes of the Advertisements. It says that the first class comprises members of the public with a general interest in the property market. It argues that they would be expected to read at least the front page of each newspaper advertisement and scan some or all of the material on the other pages. Domain contends that the second class comprises current and prospective property vendors, purchasers, lessors or lessees (property transactors) and real estate agents. It says that, having regard to their greater interest in the property market, persons in this class would scrutinise the claims made in the Advertisements more carefully, would view the claims made with an analytical mind and would not easily have been affected by the advertising claims made.

36    I take a slightly different approach. I agree that the Advertisements were targeted at a broad audience of persons with an interest in the property market, but I consider it appropriate to break the audience into the following two classes:

(a)    a first target class comprising members of the public with an interest in the property market within NSW, Victoria and Queensland. The Advertisements were intended to motivate such persons to download and use the Domain app, and to motivate property transactors to advertise their properties on the Domain app. This class will include people with varying levels of interest in property, including property transactors with a high level of interest, people considering entering the property market with a lower level of interest, and people who merely like to watch the property market and whose interest is occasional and lower again. Their level of interest may change over time. For example, a person may scour property apps and websites over a period of time looking for a property to buy or lease. Once he or she secures a property that same person is likely to have a lower level of interest in the property market and pay less attention to an advertisement about a property listings app; and

(b)    a second target class comprising real estate agents who sell or lease properties in the geographical areas covered by the advertising campaign. The Advertisements were intended to motivate them to advertise the properties of their clients on the Domain app and website (or at least to recommend to their vendors and lessors clients that they be instructed to do so), and to download and use the Domain app. I see this class as distinct from the first class because they make their living by charging commission on properties they sell or lease, and property listings apps and websites are central to their business. They will have a higher and more sustained level of interest in the property market, and a greater level of knowledge about property listings apps and websites and about the marketing techniques used in their promotion.

37    Accordingly, the case falls to be decided having regard to the representations conveyed by the Advertisements to the hypothetical ordinary or reasonable:

(a)    member of the public with an interest in the property market in NSW, Victoria or Queensland; and/or

(b)    real estate agent who sells or leases property in NSW, Victoria or Queensland.

Previous comparative advertising

38    Domain put on evidence of various examples of comparative advertising published by REA in the period prior to the publication of the impugned Advertisements, which REA did not deny publishing. The comparative advertising included that:

(a)    in December 2015 REA published a full-page advertisement in the Glen Eira/Port Phillip Leader newspaper featuring the claim that REA receives “Twice the visits of any property site”. At the base of the advertisement in fine print a footnote said “Source: Nielsen, 12 month average to October 2015, web and msite.” The footnote did not refer to app usage. REA published a similar claim in the Wentworth Courier newspaper on 27 January 2016 accompanied by a footnote that read “Source: Nielsen Market Intelligent 12 Months to Dec ‘15”. REA published similar claims in billboard advertising in Melbourne and Sydney between 1 December 2015 and 27 January 2016. Domain argues that this advertising conveyed the impression that REA had double the overall visits than Domain across all of its platforms including its app. Domain says that this was misleading because REA did not have double the visits to all its platforms, and because Domain claims to lead REA in app usage;

(b)    in February 2016 REA published an Investor and Analyst Presentation for the half year ended 31 December 2015 which contained the claim “More than twice the visits of #2 site”. An accompanying footnote, in small but visible font, qualified the claim by stating that it was based on data that excluded app visits. Domain argues that this is a significant qualifier because on any view app sessions constitute a significant component of REA’s and Domain’s audience. By way of explanation, an app visit or session is recorded each time an app is launched. The session comprises the user’s interactions with the app within a given timeframe. Each session expires after 30 minutes of inactivity, at midnight of each day, or when the app is closed. A single user can therefore record multiple sessions on any given day. The evidence indicates that when app sessions are factored into each party’s audience figures REA does not in fact receive “twice the visits” of Domain;

(c)    REA’s same half-year Investor and Analyst Presentation also contained the claim that the REA app was the “Most popular property app”. A small accompanying footnote informed the reader that the claim was based on a privately commissioned survey of 2,205 people. Domain argues that this is misleading because it is contrary to the ratings given by a far larger number of people in the All Versions Rating of the Domain and REA apps in the Apple App Store and the Google Play Store;

(d)    around February 2016 REA published a Facebook advertisement which claimed that REA’s app is “Australia’s most used property app” which appeared to be based on the fact that the REA app has been downloaded 5 million times. Domain argues that this is misleading because the number of downloads does not provide an accurate guide to the level of use, that the Domain App was downloaded more often than the REA app in 2015, and that there were more app sessions on the Domain App than the REA app for the period July 2015 to January 2016; and

(e)    also around February 2016 REA published a LinkedIn advertisement which claimed that REA had “Australia’s No.1 residential and commercial property websites”.

39    Based on this evidence, Domain contends that the property listings market is characterised by intense competition in which the parties disseminate ambit claims about the supremacy of their respective services, supported by selective reference to the available data and the subtle use of footnotes and statements to qualify, explain or contextualise such claims. It submits that such behaviour means that many persons within the target audience are likely to be acutely accustomed to (and sceptical of) promotional claims in relation to the competing apps and websites.

40    Domain did not, however, file a cross claim alleging that REA made false or misleading representations or that its conduct in publishing these claims was misleading or deceptive. As a result there was no requirement for REA to fully engage with Domain’s contentions in this regard and it did not do so.

41    I accept that REA published this comparative advertising, and some of Domain’s claims in this regard appear well founded. However, this case was not the occasion to decide whether REA’s conduct was misleading or deceptive. It is appropriate to take this comparative advertising into account in terms of the knowledge to be imputed to the target audience but it does not carry the weight for which Domain contends. Largely this is because I am not satisfied that members of the public with only a general interest in the property market would have been aware of this advertising or given it the attention for which Domain contends.

The knowledge to be imputed to the target audience

The class of members of the public with an interest in the property market

42    As I have said, this class encompasses a broad cross-section of the public with varying levels of interest in the property market, and their level of interest will fluctuate over time. The class includes people with many different characteristics including those who are knowledgeable about the property market and property apps and websites and those who are not, the shrewd, the ingenuous, those who are sceptical and read footnotes and the small print and those who do not.

43    It is appropriate to assume that members of the public with an interest in the property market would approach the Advertisements armed with a general knowledge of worldly affairs having regard to the nature of the market in which the parties operate and to the manner in which goods and services are commonly sold within that market (Country Road Clothing Pty Ltd v Najee Nominees Pty Ltd (1991) 20 IPR 419 at 422; [1991] FCA 119 at [15] (Country Road) (Heerey J); Coles at [41]). Members of this class will know that advertisers sometimes overstate the benefits of their products and services.

44    Current and prospective property transactors, who are likely to have a high level of interest in the property market, may have paid close attention to the parties’ earlier comparative marketing claims. They may have formed the view that there was a level of overstatement in the parties’ earlier advertising such that they will have approached the present Advertisements with a heightened level of caution or scepticism. Even so, a significant proportion of this class, comprising members of the public with a lower level of interest in the property market, are likely to be unaware of or will have paid little attention to the parties’ earlier marketing claims.

45    It must be kept in mind that the parties’ apps are free to download and use, and their websites are free to use. The ordinary or reasonable member of the public with an interest in the property market is unlikely to give careful attention to the claims in the Advertisements when he or she will know that if the advertised app is not as good as represented he or she can simply stop using it without suffering any detriment. He or she will be accustomed to advertising puffery and will bring the usual scepticism of the public to such claims, but I do not accept that it is appropriate to impute to him or her the heightened level of scepticism for which Domain contends.

46    Finally, in my view it is appropriate to impute to members of the public with an interest in the property market (and to real estate agents) knowledge about apps, including how they are updated and how they are rated. Apps are ubiquitous in modern life and the ordinary or reasonable member of the public or real estate agent will have downloaded and used apps. He or she would know that they usually carry a star-based consumer rating (sometimes accompanied by consumer reviews) and that well-used apps are regularly updated. The consumer rating, and the date and version number of the app appears on the download page in the relevant app store. App users are notified when the latest version is available to download.

47    The ordinary or reasonable member of either target class will know that consumer ratings of apps are only a guide, and carry little assurance of reliability. One might ask: Who has not used an app which has a high rating yet is difficult to use or contains ‘bugs’ and does not reflect its high consumer rating? Similarly, who has taken up a hotel booking or holiday accommodation on a booking app like Wotif, Tripadvisor, Booking.com or Trivago, in part based on good consumer ratings and reviews, and been quite unsatisfied with the accommodation ultimately received?

The class of real estate agents

48    Real estate agents make their living largely from the commissions they charge on properties they sell and lease. Property apps and websites are an important, probably now the most important, method by which they advertise such properties for sale or lease to prospective purchasers and lessees, and also to attract property listings from potential vendors and lessors.

49    The importance of property apps and websites to real estate agents means they are likely to have a keen interest in and knowledge of the competing platforms. Amongst other things, they will be interested in which parties’ platforms are used most often, the number and type of properties listed on them, their ease of use and their effectiveness in generating sales or leasing enquiries. They will advise their clients to advertise their properties on the most effective platforms and will use those platforms themselves. As they are likely to want information about all listed properties in their geographical area of operation, and because both apps and websites are free to use, it seems likely that many real estate agents will download and use both apps and use both websites.

50    In my opinion many real estate agents are likely to have a view, unrelated to the Advertisements, as to which property listings app or website is the best, based on their personal experience in using those platforms.

51    The evidence shows that REA and Domain understand the central role played by real estate agents in the property listings market and they have previously marketed their platforms directly at real estate agents. The real estate agents’ general business experience, their experience in property marketing, their awareness of the comparative marketing claims of Domain and REA, their own views as to the respective merits of the apps, and that they are likely to be well-versed in making puffed up advertising claims, means that it is appropriate to the class of real estate agents with a high level of knowledge in relation to property listing apps and websites and the advertising techniques used in marketing such platforms, such that they would approach the claims in the Advertisements with a high level of caution or scepticism.

E.    THE ADVERTISING CAMPAIGNS

The first phase of the campaign

52    Domain commenced the first phase of the advertising campaign by publishing the Second, Third and Fifth Advertisements on 11 February 2016. The Second and Third Advertisements were “wrap-around” advertisements in The Sydney Morning Herald and The Age newspapers, and the Fifth Advertisement comprised billboards displayed in Sydney. These Advertisements made claims that Domain has the “#1 property app in Australia”, and that “the most property listings in Sydney are on Domain” or that “the best property listings in Melbourne are on Domain” (the impugned claims).

REA commences proceedings

53    On 12 February 2016 REA commenced this proceeding and made an application for urgent interlocutory relief. On that day Domain gave an undertaking not to publish or cause to be published the impugned claims, or claims which convey the same or similar meanings in relation to the Domain app or website. That undertaking has remained on foot pending the hearing and determination of the proceeding.

54    It appears that for a short time following the undertaking Domain continued to publish one or more Advertisements which contained an impugned claim. Domain says this was a result of some miscommunication and it submits that it removed the offending Advertisements as soon as it was made aware of the publication. On 15 February 2016 REA filed a charge alleging contempt of court by the publication on 13 February 2016 of an advertisement in The Sydney Morning Herald which contained one of the impugned claims. The allegation of contempt has not yet been heard.

The second phase of the campaign

55    On 21 February 2016 Domain commenced the second phase of the campaign by publishing the Fourth, Sixth and Seventh Advertisements. The Fourth Advertisement took up the front page of The Sunday Age newspaper, the Sixth Advertisement comprised billboards located in NSW, Victoria and Queensland and the Seventh Advertisement was an advertisement on Facebook. Relevantly, these Advertisements made claims that Domain has “Australia’s highest-rated property app”.

F.    THE FIRST Phase: THE SECOND, THIRD AND FIFTH ADVERTISEMENTS

Publication of the Advertisements

The Second Advertisement

56    The Second Advertisement comprised a four page wrap-around advertisement published with the 11 February 2016 edition of The Sydney Morning Herald. REA’s complaint centres on the front page of the wrap-around, an image of which appears below:

57    The front page had the usual newspaper header at the top, below which featured a prominent heading in white font on a green background which said “#1 property app in Australia” (the No.1 Property App Claim). Below that a prominent subheading in black font appeared which said “The most property listings in Sydney are on Domain” (the Most Listings Claim).

58    Below these headings an image of a mobile telephone appeared with the Domain app displayed on its screen. Immediately to the right of the mobile telephone the advertisement said “Download our Property App today”, above logos for the Apple App Store and the Google Play Store (at which the reader could download the app). The Domain logo was prominently displayed in the bottom right corner in green font on a white background.

59    Not apparent in the image above, and in my view barely perceptible on the Advertisement itself, a small caret (i.e. a small inverted “v”) appeared at the end of the heading “#1 property app in Australia”. Domain argues that the caret indicated the existence of a footnote which qualified the claims made on the front page. The corresponding footnote was at the base of the second page.

60    The second page of the Advertisement contained large images of a mobile telephone and two “smart” watches which displayed the Domain app on their screens, alongside text describing some of the features of the app. At the bottom of that page the footnote in small font said:

As voted by consumers, All Versions rating via App Store and Google Play Store, February 2016. Manual listing count of established homes Domain versus realestate.com.au Sept 2015 to Jan 2016.

Domain says that the first sentence of the footnote was directed at qualifying the No.1 Property App Claim and the second sentence was directed at qualifying the Most Listings Claim.

The Third Advertisement

61    The Third Advertisement also comprised a four page wrap-around, this time published with the 11 February 2016 edition of The Age newspaper. REA’s complaint centres upon the front page of the wrap-around, an image of which appears below:

62    Like the Second Advertisement, the front page had the usual newspaper header at the top, below which featured a prominent heading in white font on a green background, which made the No.1 Property App Claim. Below that a prominent subheading in black font appeared which said “the best property listings in Melbourne are on Domain” (the Best Listings Claim).

63    Again, a barely visible caret was appended to the end of the headline No.1 Property App Claim and the relevant footnote appeared at the base of the second page of the Advertisement.

64    The remainder of the first and second pages of the Third Advertisement were in all material respects, except one, the same as the Second Advertisement. The only difference was in the footnote at the bottom of the page which said in small font:

As voted by consumers, All Versions rating via App Store and Google Play Store, February 2016.

Domain says the footnote qualified the No.1 Property App Claim and did not relate to the Best Listings Claim.

The Fifth Advertisement

65    The Fifth Advertisement comprised a series of billboards located in Sydney from 11 February 2016, an image of which appears below:

66    The billboards featured a prominent heading that made the Most Listings Claim. That claim sat above a subheading which said that “Sydney’s best property app just gets better”, alongside an image of a mobile telephone with the Domain app displayed on its screen, and the prominent Domain logo.

67    The billboard included a footnote in much smaller font at the bottom left which read:

Listing counts of established homes Domain versus realestate.com.au Sept 2015 to Jan 2016.

Domain says that the footnote was directed at qualifying the Most Listings Claim.

The alleged representations

68    REA alleges that the Second, Third and Fifth Advertisements conveyed one or more of the following representations to the target audience:

(a)    that the Domain app is the “#1 property app in Australia” because it allows users to view “the most property listings in Sydney” (Second Advertisement) or “the best property listings in Melbourne” (Third Advertisement) (the Combined Representation);

(b)    that the Domain app and website have more property listings in Sydney (Second and Fifth Advertisements) or Melbourne (Third Advertisement) than any other comparable app and website (the Most Listings Representation);

(c)    that the Domain app is the most widely or commonly used app in Australia (Second and Third Advertisements) (the Most Used Representation); and/or

(d)    that the Domain app has been ranked as the “#1 property app in Australia” by an independent, objective and reliable source (Second and Third Advertisements) (the Independently Ranked Representation).

Domain’s broad contentions

69    Domain denies each of the alleged representations. At a broad level it argues that they are the product of complicated or over intellectualised analysis which borders on the extreme or fanciful, and casts the ordinary and reasonable reader of the Advertisements as someone who is careless and absent-minded as to the claims made, but also someone who gives rigorous and intellectualised analysis to advertising about a free property listing database. It submits that the ordinary and reasonable members of each target class, armed with their general knowledge of worldly affairs and having regard to the nature of the market in which the parties operate, the manner in which services are commonly sold within that market, and the marketing techniques commonly deployed by both parties in that market would:

(a)    not have understood the publications to convey the representations for which REA contends;

(b)    have approached each of the publications with scepticism having regard to the laudatory claim nature of the claims made; and

(c)    not have been led into error.

Domain argues that at their highest the claims in the Advertisements may have caused ordinary or reasonable members of the target audience some uncertainty, but such uncertainty is insufficient to support a finding of misleading or deceptive conduct.

70    Domain contends that the Court should consider whether the Advertisements constitute misleading or deceptive conduct through the lens of its intense competition with REA for supremacy in the Melbourne and Sydney property listings market, notably the competition around their respective apps, which are the biggest growth area for property listing advertisements. It notes that a finding that a reputable publisher has engaged in misleading or deceptive conduct is no light matter (Country Road at 424). It contends that the Court should not suppress what it says is effective advertising in a vigorous marketplace, which the ACL is intended to promote.

71    Domain also argues that, read as a whole, the four pages of the newspaper Advertisements show that they extol the virtues and functionality of the Domain app, and not the Domain website. It says that none of the Advertisements convey representations concerning both the Domain app and website as REA alleges.

72    I note that, having regard to the greater level of knowledge and the heightened level of scepticism with which real estate agents are likely to have approached the Advertisements, REA’s easier case is to establish that these representations were conveyed to the class of members of the public with an interest in the property market. Generally speaking, if REA cannot establish that a particular representation was conveyed to the ordinary or reasonable member of that class it is unlikely to be able to establish that the representation was conveyed to real estate agents.

Whether the alleged representations were conveyed by the Second, Third or Fifth Advertisements

The Combined Representation

73    REA alleges that it was available to the ordinary or reasonable reader to understand the words “#1 property app in Australia” in combination with the words “the most property listings in Sydney are on Domain” (in the Second Advertisement) and the words “the best property listings in Melbourne are on Domain” (in the Third Advertisement). It therefore contends that the Second and Third Advertisements conveyed the representation that Domain has the “#1 property app in Australia” because it allows users to view “the most property listings in Sydney” or “the best property listings in Melbourne” (i.e. the Combined Representation).

74    Against this, Domain argues that the Combined Representation is not a natural or ordinary construction of the Advertisements. It notes the following differences between the claim that Domain has the number one property app in Australia and the Most Listings and Best Listings Claims:

(a)    the claim that Domain has the number one property app appears in large capital letters near the top of the first page of each newspaper Advertisement, in white print against a green background, whereas the Most Listings Claim and the Best Listings Claim are in smaller font, in black text rather than white;

(b)    the Most Listings Claim and the Best Listings Claim appear underneath rather than alongside the claim to have the number one property app;

(c)    the Most Listings Claim and the Best Listings Claim address the more limited and distinct geographic areas of Sydney or Melbourne while the claim to have the number one property app relates to Australia; and

(d)    the Most Listings Claim is a quantitative statement while the claim to have the number one property app is a qualitative statement.

Domain contends that these differences would signal to the ordinary or reasonable reader that the Most Listings Claim and the Best Listings Claim are subsidiary and distinct claims to the claim to have the number one property app, and are not to be read in combination.

75    Domain argues that the evidence of Mr Rechtman, REA’s Executive General Manager, Residential, neatly encapsulates the proper construction of the Second and Third Advertisements as they would be understood by the ordinary reader. Mr Rechtman described the Second Advertisement as containing a primary claim that the Domain app is the “#1 Property App in Australia” and a secondary claim that “the most property listings in Sydney are on Domain”, a distinction which Domain says clearly marks the two claims out as separate and not to be read in combination.

76    When dealing with alleged the No.1 Property App Claim, Domain contends that it and the Most Listings Claim are qualified by footnotes which explain the claims. It argues that:

(a)    the claim that Domain has the “#1 property app in Australia” is a qualitative claim based on the qualities and functionality of the Domain app (as referred to in the Advertisements) supported by reference to the All Versions Ratings assigned to the app by consumers on the Apple App Store and the Google Play Store as the footnote states; and

(b)    the claim that “the most property listings in Sydney are on Domain” is based on the number of listings of established homes in Sydney on the Domain app compared to the REA app as the footnote states.

77    Domain argues that, at least, property transactors and real estate agents (who would have read the footnote) would have understood the claims in that way. It says that ordinary consumers, sophisticated property investors and real estate agents alike will be aware from their own worldly experience that laudatory claims of this kind, qualified by footnotes or other disclaimers, are commonly made. In my view, if Domain was correct in this contention (which it advanced in another context) then it would also follow that the Combined Representation was not conveyed.

78    Domain also argues that, even if the No.1 Property App Claim is understood in combination with the Best Listings Claim (which it describes as “quintessential puffery”) it would be understood as just a laudatory claim which does not convey any definite meaning.

79    As I have said, an advertisement targeted at a broad audience as in the present case may be found to convey a representation if the meaning is reasonably open, in the sense of being one of several available meanings, to the ordinary or reasonable member of that audience: Tobacco Institute at 50; Coles at [47]. In construing the available meanings it must be kept in mind that the ordinary or reasonable member of each of the target classes I have identified will have read the Advertisements fleetingly and will not have closely studied or carefully analysed their constituent parts, particularly when the Advertisements are about a property listings app which is free to download and use. He or she is likely to have taken away only the broad thrust of the Advertisements: TPG at [45]; Tobacco Institute at 4; ACCC v TPG at [38], [43].

80    I take a different view in relation to the Second and Third Advertisements and it is convenient that I start with the latter Advertisement.

The Third Advertisement

81    In my view it was reasonably available for the ordinary member of each target class to read the claim in the Third Advertisement that Domain has the “#1 property app in Australia” in combination with claim that “the best property listings in Melbourne are on Domain”. While the positioning, font size and text colour of the claim to have the “#1 property app in Australia” differed from that of the other claim, having regard to the fleeting attention likely to be given to the Advertisement it was open to the ordinary reader to understand them together.

82    Indeed, reading them together is a natural reading. Domain’s contention that the claims would be read as distinct because one related to Australia and the other related to Melbourne attributes a care and attentiveness to the ordinary reader which is at odds with the reality of the way in which people treat advertising. The ordinary reader could reasonably understand the claim that Domain has the “#1 property app in Australia” as being informed by the claim that it has “the best property listings in Melbourne”.

83    To the extent that Domain advanced the argument in this context, there is little merit in the contention that the ordinary or reasonable reader would understand the claims in the Third Advertisement by reference to the qualifying information in the footnote. There was no footnote to the Best Listings Claim and (for reasons I explain in relation to the Second Advertisement) the ordinary reader would not have read or taken in the footnote to the claim that Domain has the “#1 property app in Australia”.

84    While I consider it was reasonably available to the ordinary reader to understand the Third Advertisement as making a claim that Domain has the “#1 property app in Australia” because it has “the best property listings in Melbourne”, in my view that would not convey a definite representation as to the superiority of Domain’s app upon which the ordinary reader would have relied.

85    I accept that it is a question upon which reasonable minds may differ, but in my view the broad claim that Domain has the number one property app in Australia, understood as a distinct or separate claim, is just puffery. Obviously, it makes a claim of superiority but that claim has little content. In my view it merely invites the question “No.1 in what way?”: Unilever Australia Ltd v Goodman Fielder Consumer Foods Pty Ltd [2009] FCA 1305 (Unilever) at [27] (Buchanan J).

86    The ordinary reasonable reader would be accustomed to puffing claims in advertising, and would know that claims to be “No.1”, “the best” or “the greatest” are commonly made in trade. The question as to what makes one property app better than another, or “No.1”, depends on so many variable subjective or objective factors that it does not carry a definite meaning. Having regard to their worldly knowledge and experience, such an indefinite claim as to the superiority of Domain’s app over its competitors is, without more, insufficient to convey a definite representation upon which the ordinary or reasonable reader would rely. It would be understood as an advertiser’s flourish or merely as reflecting Domain’s own self-serving commendation of its app. This is particularly clear in relation to the class comprising estate agents who would have been exposed to the parties’ earlier marketing efforts, would have a greater knowledge of the subtleties of their advertising techniques, and would be well-versed in exuberant or overstated marketing claims.

87    When the broad claim that Domain has the number one property app in Australia is understood in combination with the claim that it has the best property listings in Melbourne, the representation conveyed is even less definite. I say this because it is unclear what meaning the claim to have the best property listings in Melbourne would convey to the ordinary or reasonable reader. For example, would the claim be understood to mean that the property listings on the Domain app are the type of properties for which the reader was searching, that the property listings are properties of a higher standard, that the properties are located in better socio-economic areas, that the property listings are better presented, or that the property listings are easier to access and use? Whether some property listings are the “best” or “better” than others will depend upon many subjective and objective variables and necessarily involves a number of value judgements.

88    In my view the ordinary member of either target class would not reasonably understand the claims, in combination, to be an assertion of the superiority of Domain’s app upon which he or she could rely, but rather as just Domain’s broad self-serving commendation of its own app over those of its competitors.

89    That is not to say that puffed statements cannot convey a definite meaning upon which an ordinary or reasonable reader could rely. For example, in Byers v Dorotea Pty Ltd (1986) 69 ALR 715; (1987) ATPR 40-760 (Byers) Pincus J held that a representation that apartments advertised for sale in a new apartment tower would be “bigger and better” than the apartments in a nearby apartment tower constituted misleading or deceptive conduct. In Unilever Buchanan J held that a claim that the brand of table spread, Logicol, was “No 1 recommended” was false and misleading insofar as it represented that it was recommended No.1 for lowering cholesterol absorption (at [24]-[32]). His Honour found that the logo containing the “No.1 recommended” claim made “a sufficiently clear representation of superiority” from the competitor’s product (at [32]).

90    However the question of whether a puffing advertising claim conveys a definite representation or not will depend upon the context and setting in which the claims appear and may involve questions of degree and impression. To a significant extent the authorities turn on their own facts. For example, in Byers the applicants alleged (at 717 and 719) that they were induced to purchase apartments in a new apartment tower by representations that:

(a)    the new tower would be a superior building to the nearby apartment tower;

(b)    the apartments in the new tower would be bigger and more sumptuous;

(c)    there would be a substantial lobby on the ground floor of the new tower; and

(d)    the design, construction and finish of the new tower would be of a high standard and superior in every respect to the nearby apartment tower.

The claims that the new apartments would be “bigger and better” were specific and quantifiable and related to the size of the apartments, the provision of a larger lobby, and the quality of the finishes.

91    Similarly, in Unilever the product packaging contained the claim “No.1 recommended for dietary change” above a prominent subheading which said “Lowers Cholesterol absorption”, made in close proximity to the Heart Foundation logo which signified the Heart Foundation’s approval of that product. Given that context it is unsurprising that Buchanan J reached the view that the “No.1 recommended” claim was likely to convey the false impression that the product was No.1 recommended for lowering cholesterol absorption (at [26]) and created a real chance that potential purchasers would think the recommendation came from the Heart Foundation (at [28]).

92    Both cases involved a level of specificity which goes further than the indefinite claim that Domain has the number one property app in Australia because it allows users to view “the best property listings” (whatever that means) in Melbourne.

93    Numerous authorities recognise that a broad claim as to the superiority of a particular product or service may be insufficient to convey a definite representation upon which the ordinary or reasonable reader would rely. For example, in Collier Constructions Pty Ltd v Foskett Pty Ltd (1990) 19 IPR 44 at 54 Gummow J described the comparative advertising claim that Collier Constructions offered the “best deal” in home construction as “eulogistic”. His Honour said that the claim would be understood by any reasonable consumer as conveying no more a precise meaning than that the advertiser regarded itself as offering the best deal, and would not be taken as a representation as to the superiority of that which was offered over what was offered by its competitors upon which the consumer might rely. In HCF (at 491), Morling J (with whom Fox and Davies JJ agreed) said that HCF’s claim to offer the “best value” health care cover depended upon so many variable subjective and objective factors that it did not carry a definite meaning.

94    REA contends that the Combined Representation in this Advertisement conveys a definite representation on the basis that the claim that Domain has the “#1 one property app in Australia” suggests that there had been a comparison and ranking of property listing apps and that the most obvious and natural basis for comparison is the number of property listings. It argues that the ordinary or reasonable reader will therefore understand the claims in combination to mean that Domain has the number one property app because it has the most property listings in Melbourne. I do not accept this. It is not a natural reading to construe “best property listings” as “most property listings”, and I do not consider that the ordinary or reasonable reader would understand “best” property listings in that way. It is unlikely that the ordinary reader would engage in the reasoning process for which REA contends given the brief attention he or she would give the Advertisement.

95    REA further argues that even if the Combined Representation is devoid of meaning it should nevertheless be held to be misleading or deceptive. In this regard it relies on Lindgren J’s observations in Eveready Australia Pty Ltd v Gillette Australia Pty Ltd (No 4) [1999] FCA 1824 (Eveready) (at [59]) where his Honour said:

If I had concluded that on close analysis the commercials were devoid of meaning I would have held them to be misleading and deceptive nonetheless.  The reason why I would have done so is that the words “lasts up to four times longer” hold out to the viewer that Duracell batteries do offer a meaningful advantage over competing brands.  The specificity of those words is inconsistent with the nature of advertising puff.  To my mind it is misleading and deceptive to appear to a reasonable television viewer to be saying something meaningful on which the viewer is invited to rely as a basis for action, when only a degree of analysis foreign to the realities of the market place would show the viewer that what was said is meaningless.

It also relies on the remarks of Buchanan J in Unilever where his Honour said that, even if the “No.1 recommended” claim in that case was devoid of meaning, it would still be misleading or deceptive because some claim for the qualities, benefits or reputation of the relevant product was being made (at [31]).

96    This issue involves questions of impression and degree and I take a different view in the circumstances of the present case. In Eveready Lindgren J considered the specificity of the claim that Duracell batteries lasted “up to four times longer” marked that representation out as more than mere puffery and the advertisement said something meaningful to the consumer. The same can be said of the claims in Unilever where the “No.1 recommended” claim appeared in close proximity to a claim to lower cholesterol absorption and to National Heart Foundation approval. That is not true of the subject Advertisement where the broad laudatory claim to have the number one property app is accompanied only by a similarly indefinite claim to have the best property listings.

The Second Advertisement

97    As with the Third Advertisement, I consider it was reasonably available for the ordinary member of each target class to read the claim that Domain has the “#1 property app in Australia” in combination with claim that “the most property listings in Sydney are on Domain”. Again, while the positioning, font size and text colour of the claim to have the “#1 property app in Australia” differed from that of the claim to have the most property listings, having regard to the fleeting attention likely to be given to the Advertisement it was open to the ordinary reader to understand them together.

98    In my view reading the claims in combination is a natural reading. Domain’s contention that the claims would be read as distinct because one claim is qualitative and related to Australia and the other is qualitative and related to Sydney is at odds with the reality of the way in which people treat advertising.

99    As I said in relation to the Third Advertisement I consider that the broad claim that Domain has the “#1 property app in Australia", understood as a distinct or separate claim, is just puffery which does not convey a representation as to the superiority of the Domain app upon which the ordinary or reasonable reader would rely. It has little content, and it merely invites the question “No.1 in what way?” However, it was reasonably available to understand that question as being addressed by the specific and quantifiable claim to have the most property listings in Sydney, such that the claim that the Domain app has the most property listings in Sydney illustrated the basis for the No.1 Property App claim and gave it content. In combination, the specificity of the claim is such that the ordinary reader would understand it as a representation upon which he or she could reasonably rely.

100    When a reputable company like Domain accompanies a broad claim to have the number one property app in Australia with a specific (and quantifiable) claim to have the most property listings in Sydney, the specificity of the latter claim tends to mark the broad claim out as more than just puffery: see Eveready at [59].

101    There is no merit in the contention that the ordinary or reasonable reader would understand the claims in this Advertisement (or in the Third Advertisement) by reference to the qualifying information in the footnote. The caret which indicated the existence of the footnote was so small as to be barely perceptible and it was unlikely to be noticed by the ordinary or reasonable reader. Even if the caret was noticed it is unlikely that the reader would have thought that it signified a footnote. Qualifying footnotes to advertisements usually appear on the same page as the advertising claim, but the footnote in the newspaper Advertisement did not appear until the following page. It is unlikely that the ordinary reader would have thought the caret signified a footnote because at a glance it would have been apparent there was none.

102    Nor do I accept Domain’s contention that the claim to have the number one property app in Australia was the type of claim which would have arrested the attention of the ordinary reader such that he or she would have searched for and located the footnote, and thus understood the claims in their proper context. An advertiser’s claim to be “No.1” or “the best” is not so unusual as to arrest the reader’s attention. In my view it is quite unlikely that the ordinary or reasonable reader would have thumbed through to the second page to find the footnote. As a result the ordinary or reasonable reader would not understand the Advertisement by reference to the qualifying information in the footnote.

103    I consider the Second Advertisement conveyed the Combined Representation to the ordinary or reasonable member of each target class, being that Domain has the number one property app in Australia because it has the most property listings in Sydney. Put another way, it conveyed that representation to a not insignificant number of the target audience.

The Most Listings Representation

104    The sub-heading in the Second Advertisement and the headline in the Fifth Advertisement made prominent claims that the Domain app has “the most property listings in Sydney” (i.e. the Most Listings Claim). The subheading in the Third Advertisement made the prominent claim that the Domain app has “the best property listings in Melbourne” (i.e. the Best Listings Claim). REA alleges, on the basis of these claims would be understood as distinct claims, that they conveyed a representation that “the Domain app and website have more property listings in Sydney than any other comparable app and website” (i.e. the Most Listings Representation).

105    As I have said, the second sentence in the footnote to the Second Advertisement (in The Sydney Morning Herald) said:

Manual listing count of established homes Domain versus realestate.com.au Sept 2015 to Jan 2016.

The footnote to the Fifth Advertisement (on billboards) said:

Listing counts of established homes Domain versus realestate.com.au Sept 2015 to Jan 2016.

106    Domain says that the footnote to the Advertisements was intended to convey that there were more “established properties” (meaning homes that were complete and ready for immediate occupation) in Sydney advertised on the Domain app than on the REA or any comparable app. Ms Cruickshank, Domain’s Chief Editorial and Marketing Manager, testified that she carefully watched the number of listings of established properties on the parties’ respective platforms from September 2015 to January 2016, and that it was only upon reaching the view that Domain had leadership in listings of established properties in Sydney over that period that she decided to make the Most Listings Claim. She conceded that, while the count was close, she knew that REA was ahead in terms of property listings overall in Sydney at that time.

107    Domain argues the claim that Domain has the “#1 property app in Australia” is the type of claim that would have arrested the attention of the ordinary or reasonable reader and caused, at least, property transactors and real estate agents to search out the footnote. It says they would therefore have understood the qualification to the Most Listings Claim. It argues that, although small, the caret in the Advertisement was readily discernible in white print against a green background and was “there to be read“: Coles at [41]. Once read, it argues that the footnote qualifies the claim. Domain however accepts that members of the public with a lower level of interest in the property market are unlikely to have read the footnote, and that it is appropriate to break the audience of these Advertisements into two groups: (a) those who read both the primary claims and the qualifying information in the footnote; and (b) those who read only the primary claims.

108    Importantly, Domain concedes that members of the public who did not read the footnote would not have understood the claim to have the most listings in Sydney as being limited to established properties. It accepts that such readers may have understood the Second and Fifth Advertisements to convey the representation that the Domain app has more property listings in Sydney than any other comparable property listing app. It accepts that such a representation is identical to the Most Listings Representation alleged by REA, except that Domain’s posited representation is limited in terms to property listings on its app, rather than relating to listings on both its app and website. There is little significance to that difference when it is common ground between the parties that the property listings on the Domain app and website were the same (as were the listings on the REA app and website).

109    In relation to this representation I take a different view about the Second and Fifth Advertisements to the view I take about the Third Advertisement.

The Second and Fifth Advertisements

110    I consider it plain that the Second and Fifth Advertisements conveyed the Most Listings Representation to the ordinary or reasonable member of each target class.

111    I say this, first, because these Advertisements prominently made a primary claim that the Domain app has the most property listings in Sydney. On a natural reading, those claims were not limited to established properties. The ordinary or reasonable reader, reading the Advertisements fleetingly would have taken away only the thrust of those primary claims.

112    Second, for the same reasons as I set out at [101]-[102] above the footnote in the Second Advertisement was insufficiently prominent to qualify the primary message.

113    In the Fifth Advertisement, the footnote on the billboard was in small font and could only be read and taken in to the extent that the circumstances allowed. The evidence does not disclose the exact location of these billboards but a later billboard was situated next to a busy road intersection and it is likely that the relevant billboards were similarly situated in areas with a high level of passing car and/or foot traffic. Many persons would have only fleetingly read the billboards because their focus was on the surrounding road or pedestrian traffic, or they would have seen the billboards at a distance from a moving car and their view may have been compromised or obstructed. They would have only taken in the primary message: see Australian Competition and Consumer Commission v Singtel Optus Pty Ltd [2010] FCA 1177 at [5] (Perram J); ACCC v TPG at [37] and [95]-[96].

114    Third, this is not a case where the alleged representation was conveyed because the target audience was disposed, independently of Domain’s conduct, to give close attention to some words and ignore the balance of the Advertisement. In the newspaper Advertisement Domain chose to obscure the qualifying information in the footnote by using a barely perceptible caret and by placing the footnote on a different page. In the billboards it chose to put the qualifying information in small font in a footnote which was unlikely to be noticed or taken in because of the circumstances in which the billboard would be viewed. I infer that Domain intended to convey the impression that its app had the most property listings in Sydney rather than limit the representation to established properties only. Making the broader claim was to Domain’s commercial benefit and the footnote undercut that claim. In such circumstances it may more readily be inferred that the Most Listings Representation was in fact conveyed by this advertisement: TPG at [55].

115    Fourth, Ms Cruickshank, Domain’s Chief Editorial and Marketing Officer, sought to explain the Most Listings Claim on the basis that Domain considered that most members of the public with an interest in the property market were concerned with the number of properties that are advertised and ready for immediate sale and possession (which she described as “established properties”). She said that Domain’s research shows that when consumers think of a property listing they are more likely to envisage an established property and that consumers frequently expressed frustration through surveys when searches of property listings turned up properties which had not yet been built or were subject to approvals.

116    The evidence shows that the great majority of property listings on the parties’ competing platforms are established properties, and I accept that most consumers will be interested in such properties rather than incomplete or new developments. Even so, many prospective property transactors were likely to have been interested in purchasing other property such as “off the plan” apartments, or a new house and land package in a housing estate which had not yet been built. I accept the evidence of Mr Ruiz, REA’s Chief Digital Officer, that many consumers value the capacity of a property listings app or website to provide a comprehensive picture of all properties available for sale or rent in an area at any one time. Such persons are unlikely to have understood the Most Listings Claim as being limited to established properties only and they will have made up a not insignificant number of the class.

117    In my view the Second and Fifth Advertisements conveyed the Most Listings Representation to the ordinary or reasonable member of each target class. Put another way they conveyed that representation to a not insignificant number of the target audience.

The Third Advertisement

118    REA’s submissions in this regard are not clear but it appears to contend that the expression “the best property listings in Melbourne are on Domain” in the Third Advertisement, understood as a distinct claim, conveyed the Most Listings Representation. However, it said little in its submissions as to why the ordinary or reasonable reader (in either target class) would have understood the claim in that way.

119    I do not accept this. On a natural reading, the claim that Domain has the best property listings in Melbourne would not be understood by the ordinary or reasonable member of either target class to mean that the Domain app and website have more property listings. That meaning is not reasonably available in my view. In any event, for the reasons previously set out, I consider the ordinary or reasonable reader would have understood the Best Listings Claim to be just puffery and not a representation as to the superiority of the Domain app upon which he or she could rely.

The Most Used Representation

120    REA contends that, understood as a distinct claim, the expression “#1 property app in Australia” (i.e. the No.1 Property App Claim) in the Second and Third Advertisements suggests that there is some objective measure by which property apps can be ranked, and that the Domain app is the highest-ranked property app according to that measure. It argues that, having regard to the importance of exposure and usage for apps which advertise property listings, it conveyed the meaning that “the Domain app is the most widely or commonly used property app in Australia” (i.e. the Most Used Representation). REA argues that this meaning is particularly likely to have been conveyed to real estate agents and current or prospective property transactors who it contends are accustomed to seeing promotional material which makes such claims based on usage metrics and who make advertising decisions based on which service is likely to reach the greatest audience.

121    I do not agree. First, although I accept that it is appropriate to apply a more critical eye to comparative advertising, in my view the ordinary and reasonable member of each target class would have understood the broad and non-specific No.1 Property App Claim to be just puffery. As I have said, the ordinary reader would be accustomed to puffing claims in advertising, and would know that claims to be “No.1”, “the best” or “the greatest” are commonly made in trade. The claim has little content and it is likely to be understood as just Domain’s self-serving commendation of its own app over that of its competitors, and not a definite representation of superiority upon which the ordinary or reasonable reader would rely. Contrary to REA’s contention, although real estate agents may be accustomed to seeing promotional material based on usage metrics, they must be treated as approaching such a claim with a high level of scepticism. They would not understand the claim as conveying the alleged representation.

122    REA argues that the existence of the footnotes shows that Domain did not intend the No.1 Property App Claim to be understood as just puffery. That may or may not be so when, as I said (at [114]), I would infer that Domain chose to obscure the footnotes. In any event the central question is whether the Advertisements conveyed the Most Used Representation, not whether Domain intended them to do so.

123    Second, even if I am wrong in treating the No.1 Property App Claim as puffery, it should be kept in mind that the Advertisements made no claim that the Domain app was the most commonly or widely used app in Australia. REA contends that the representation is to be implied and its contention involves accepting that the ordinary or reasonable reader understands that; (a) consumer usage of a property listings app is a significant metric within a property listings business; (b) the consumer usage of a property listings app operated by a competitor is capable of being accurately measured; (c) Domain was in a position to accurately measure the consumer usage of its own and REA’s app; and (d) Domain had undertaken measurements such that it was in a position to legitimately make the Most Used Representation. The authorities warn that it is inappropriate to approach advertising claims on such a basis: see ACCC v Telstra at [49].

124    It is unlikely that the ordinary member of the public with an interest in the property market would possess such knowledge or have such an understanding. Some real estate agents may have the relevant knowledge but they will have approached the No.1 Property App Claim with a high level of caution or scepticism such that they are likely to have understood it as puffery. Having regard to the fleeting attention likely to be given to the Advertisements it is unlikely that the ordinary reasonable member of either target class would have given the Advertisements the suggested level of attention and analysis, or that he or she would have engaged in the suggested reasoning process.

The Independently Ranked Representation

125    In the alternative REA contends that the No.1 Property App Claim, understood as a distinct claim, suggests that there has been a ranking of property apps by some external source, and that Domain is seeking to promote the fact that the Domain app has been ranked by that source as the number one property app in Australia. It argues that, because of the prominent and unqualified way in which the No.1 Property App Claim is made, the ordinary or reasonable member of each target class was likely to assume that such a claim could only (legitimately) have been made with a firm evidentiary foundation, and that therefore such a ranking must have been carried out by a source which is independent, objective and reliable (i.e. the Independently Ranked Representation).

126    I do not consider that the Advertisements in the first advertising campaign conveyed the Independently Ranked Representation to the ordinary reasonable member of either target class.

127    First, as I have said, in my view the ordinary or reasonable member of each target class would understand the No.1 Property App Claim as just puffery and not as making a representation as to the superiority of Domain’s app upon which he or she could reasonably rely.

128    Second, even if I am wrong in treating the No.1 Property App Claim as just puffery, REA’s contention that the Independently Ranked Representation was conveyed involves treating the ordinary or reasonable reader as having the understanding that; (a) there are criteria by which property apps can be ranked; (b) an external body exists which ranks competing apps by such criteria (c) information (including information which is likely to be confidential) regarding property apps is available to the external body; (d) using that information, the external body had ranked the property apps; and (e) Domain was promoting its app by stating that it has been ranked by that external body as the number one property app. I do not accept that the ordinary reasonable reader in either target class would have given such careful attention or analysis to the claim, or engaged in the suggested reasoning process.

129    The ordinary or reasonable member of each target class must be treated as being familiar with apps and knowing that they are rated by consumers on the relevant app store. I see it as quite unlikely that he or she would reach the view that there was some unidentified independent body which had the role of assessing property apps against some (unknown) criteria. It is uncontroversial that no such body exists, and I am not satisfied that the ordinary reader would think that it did.

Conclusion regarding the representations conveyed in the first advertising campaign

130    To recap the views expressed above, I consider that:

(a)    the Second Advertisement conveyed the Combined Representation, but the Third Advertisement did not; and

(b)    the Second and Fifth Advertisements conveyed the Most Listings Representation, but the Third Advertisement did not.

None of the Advertisements conveyed the Most Used Representation or the Independently Ranked Representation.

Whether the Combined Representation and/or the Most Listings Representation were false or misleading or constitute misleading or deceptive conduct

131    The Combined Representation (in the Second Advertisement) and the Most Listings Representation (in the Second and Fifth Advertisements) conveyed the representation that the Domain app has more property listings in Sydney than any other comparable app. The question as to whether these representations were false or misleading or constitute misleading or deceptive conduct may be dealt with by reference to the evidence as to the number of property listings in Sydney on the Domain app compared to the REA app in the period preceding the Advertisements.

132    Each party put on evidence from its employees as to the number of Sydney property listings on their apps and websites. Confronted with the parties’ competing calculations and their attacks on the opposing party’s calculations, I directed the parties to appoint a joint independent expert. The parties appointed Mr Stone, a forensic accountant with KordaMentha, to review the property listings data produced by the parties and to calculate the number of property listings in Sydney on the parties’ respective platforms in the period from September 2015 to February 2016 inclusive (the relevant period). Mr Stone did so in relation to:

(a)    all property listings; and

(b)    listings of “established properties” only.

133    Mr Stone was requested to use two different calculation methods for what were described as “parent” listings and “child” listings in relation to listings of apartments within apartment complexes. Under a parent listing an apartment complex that has multiple apartments of different types is treated as a single property listing. Under a child listing each apartment category within an apartment complex is treated as a separate property listing. By way of explanation where a property listing comprised a parent listing and one or more child listings:

(a)    under the first method of calculation, Mr Stone was requested to only count the parent listing. On this basis, for example, a Sydney apartment tower would be measured as one property listing (the parent) for the entire building;

(b)    under the second method of calculation, Mr Stone was requested to only count the child listing. On this basis only separate listings for each category of apartment within the apartment tower, such as one bedroom, two bedrooms, three bedrooms and penthouse, would be measured.

134    For the relevant period Mr Stone calculated:

(a)    the total property listings in the Greater Sydney Area; and

(b)    the total listings of established properties (as defined) in the Greater Sydney Area;

on the Domain and REA apps and websites. He did so on both the first and second methods of calculation, and by reference to:

the number of property listings on Domain Check Days (being the nominated day in each month of the relevant period which Domain had used for its check of property listings);

the average number of property listings on Domain Check Days;

the total number of unique property listings for each month of the relevant period; and

the average daily number of property listings for each month of the relevant period.

The trial proceeded on the basis of Mr Stone’s calculations and the parties did not rely upon their employees’ calculations.

135    Mr Stone’s evidence establishes that, although the difference in the property listings on the parties’ respective platforms was not great, over the relevant period REA had more property listings overall in the Greater Sydney Area than Domain. Although Domain’s platforms had more total unique listings than REA’s platforms in September 2015 (when measured using the first and second methods) and in November 2015 (when measured using the first method only) REA’s platforms had more listings in those months measured by the total number of listings on Domain check days and the average number of listings per day. In November 2015, the total number of unique listings on the parties’ platform was the same when calculated on the second method for parent and child listings.

136    Therefore (except for November 2015 when the Domain platforms had more listings using the first method of calculation only) from October 2015 to February 2016 the total number of unique listings on REA’s platforms was equal to or higher than the total number of unique listings on Domain’s platforms.

137    Mr Stone’s evidence shows that the Most Listings Representation was false: the Domain app and website did not have more property listings in Sydney than any other comparable app and website. It also shows the Combined Representation was false: the Domain app is not the No.1 property app in Australia because it allows users to view the most property listings in Sydney.

138    It is, however, worth noting that Mr Stone’s evidence shows that throughout the relevant period Domain’s platforms had more listings of established properties in the Greater Sydney area than REA’s platforms using both the first and second methods of calculation. His evidence also shows that, using both methods of calculation, the established properties comprised more than 90% of the total listings. It follows that, had the footnote to the Advertisements been sufficiently prominent to qualify the false primary message, the Most Listings Representation would not have been misleading.

139    Domain argues that to the extent that there is any disparity between its primary claims and the true position, that disparity is not so great as to have required the qualifying footnote in the Second and Fifth Advertisements to have appeared more prominently. It submits that the fact that established listings comprise so great a proportion of total listings provides empirical support for Ms Cruickshank’s evidence that most consumers in the Sydney property market are interested in listings for established properties and, accordingly, that listings of established properties represent Domain’s main market in Sydney.

140    Domain argues that:

(a)    there is no material disproportion between the primary claim made in the Second and Fifth Advertisements and the qualifying information in the footnotes to those Advertisements;

(b)    although the Most Listings Representation may not literally be true it was not disparate from the true position in any material or meaningful way; and

(c)    to the extent that there is any disparity between the Most Listings Representation and the true position, such disparity only comes into focus by drawing “fine and precise distinctions” of the kind warned against in Stuart Alexander.

It argues that the Court should not find that the Most Listings and Combined Representations were false or misleading or constitute misleading or deceptive conduct.

141    The authorities (set out at [27] above) show that the degree of prominence which must be given to a qualifying statement may depend on the potential of the primary statement to be misleading. Domain relies on the decision in Stuart Alexander in which the impugned advertisement claimed that a competitor’s brand of instant coffee was twice as expensive as the instant coffee being promoted, when in fact it was only about 50% more expensive. Lockhart J considered that a robust approach should be taken in relation to such “puffing” advertisements and said that it was inappropriate to make “fine and precise” distinctions in advertising of that kind. His Honour did not consider the advertisements to be misleading or deceptive: Stuart Alexander at 164-165.

142    Domain draws a distinction between advertisements which make an offer about a product or service and then materially diminish the offer in the footnote or fine print, and advertisements which make a claim about a product or service and then explain the offer or source information in the footnote or fine print. It puts the Second and Fifth Advertisements in the latter category. It contends that the Most Listings Claim did not offer consumers a bargain that was materially diminished by the qualifying information in the footnote, but instead made puffed up claims about the Domain app which were then explained or sourced in the footnote. It argues that the Court should adopt the robust approach taken in Stuart Alexander and should not conclude that the Most Listings Representation was not misleading or deceptive.

143    I do not accept Domain’s contentions in this regard. I have no difficulty accepting that there is a difference between advertising in which the primary message offers consumers an attractive deal which is materially diminished in a footnote or fine print, and advertising in which the footnote qualifies the primary claim without contradicting it. I do not, however, accept Domain’s contention that the present case falls into the latter category. Nor do I consider that the present case is analogous with Stuart Alexander. In that case the broad thrust of the advertisement - that the competitor’s brand of instant coffee was more expensive - was true, but the Court found that the competitor had puffed its claim and exaggerated the price differential. In the present case Domain’s representation that its app had the most property listings in Sydney was false unless the reader understood that it only related to established properties.

144    The evidence shows that the more people that use a property listings app or website the more others will want to advertise their properties on it. The Second and Fifth Advertisements each contained the primary message that the Domain app and website have more property listings in Sydney than any comparable app or website. This message was likely to have had some importance for many members of the target audience. The footnote undercut that specific (and quantifiable) claim by limiting it to established properties only, yet the qualifying information was given no prominence and was unlikely to be read or taken in.

145    I do not accept the contention that the distinction between total property listings and listings of established properties is “fine or precise”, as Domain contends. I accept Mr Ruiz’s evidence that the more properties that are listed on a particular app or website the more likely that prospective property purchasers or lessees will want to use that service to search for properties. While most consumers are likely to be interested in listings of established properties, many members of the target classes will be interested in downloading and using the most comprehensive app (i.e. the app with the most listings). I infer that Domain understood the significance of the distinction because it chose to obscure the qualifying information which would have made the distinction clear. That was to its commercial advantage.

146    I find that Domain’s conduct in publishing the Combined Representation in the Second Advertisement and the Most Listings Representation in the Second and Fifth Advertisements constitutes misleading or deceptive conduct or conduct which is likely to mislead or deceive in breach of s 18 of the ACL. I also find that in representing that the Domain app is the number one property app in Australia because it allows users to view the most property listings in Sydney Domain made false or misleading representation that its service was of a particular standard, quality or grade, in breach of s 29(1)(b) of the ACL.

Were the other alleged representations false or misleading and did they constitute misleading or deceptive conduct (if they were conveyed)

147    I have found that:

(a)    the Third Advertisement did not convey the Combined Representation or the Most Listings Representation; and

(b)    in the Second and Third Advertisements, the No. 1 Property App Claim did not convey the Most Used Representation or the Independently Ranked Representation.

It is unnecessary to decide whether those representations would have been false or misleading had they been conveyed, but as the matter was fully argued and having regard to the possibility of an appeal it is best that I briefly set out my views in this regard.

The Combined Representation in the Third Advertisement

148    There is no evidence to show, and Domain did not contend, that its app and website had the best property listings in Melbourne (whatever that means). To the extent that the evidence dealt with this issue it tends to show that the Domain and REA apps and websites carry similar property listings.

149    Accordingly, if (contrary to my view) the Third Advertisement conveyed the representation that the Domain app is the No.1 property app in Australia because it allows users to view the best property listings in Melbourne (and that representation had a definite meaning upon which the ordinary or reasonable reader would rely) it was false or misleading and Domain’s conduct was misleading or deceptive or likely to mislead or deceive.

The Most Listings Representation in the Third Advertisement

150    There is no evidence to show, and Domain did not argue, that its app and website had more property listings in Melbourne than the REA app and website. Accordingly, if (contrary to my view) the Third Advertisement conveyed the Most Listings Representation, it was false and misleading and Domain’s conduct was misleading or deceptive or likely to mislead or deceive.

The Most Used Representation

151    Domain denies that the No.1 Property App Claim conveyed the Most Used Representation but argues in the alternative that there was a legitimate basis for any such representation.

152    First, Domain relies on evidence which shows that in the period between January 2015 and January 2016 the Domain app recorded 167.389 million app sessions, which included a significant increase in the number of app sessions per month across that period. Domain’s evidence indicates that for each month between July and October 2015 the Domain app recorded more app sessions than the REA app, and that between July 2015 and January 2016 the Domain app led the REA app both in total app sessions and average monthly app sessions. Domain argues that this was a legitimate basis for any claim (if made) that the Domain app is the most widely or commonly used property app in Australia.

153    However, the evidence as to app usage goes both ways. Mr Ruiz and Ms Blake, REA’s Digital Analytics Manager, gave essentially consistent evidence in relation to the REA app sessions in each month from January 2015 to January 2016. Their evidence did not though accord with the evidence of Mr Rechtman for the months November 2015 to January 2016. Nor did it accord with some of the app session data set out in “Fast Fact Sheets” published by REA over the period January 2015 to November 2015. In some cases, the disparity was a few hundred sessions and in others it was much greater. Mr Ruiz explained that the disparity was due to accounting errors wrought through “propagation” and by incorrect tagging.

154    Although I was left with some questions as to the reliability of REA’s evidence in this regard, on balance I consider it shows that there were more app sessions on the REA app in each month between January and June 2015 and again in each of the three months immediately preceding the publication of the Advertisements (between November 2015 and January 2016). The evidence tends to show that the REA app also had more overall app sessions and more average number of app sessions per month than the Domain app (albeit only by a small number) between September 2015 and January 2016.

155    Domain argues that the fact that differing conclusions can be drawn from the data indicates that these are matters about which reasonable minds may differ by reference to different aspects of the metrics. That may or may not be so. It suffices to note that, if it were necessary to decide, I would not be satisfied that the app session data supports a claim that the Domain app is the most widely or commonly used property app in Australia.

156    Second, Domain relies on evidence as to the number of recent downloads of the Domain app as compared to the REA app. Each time an app is downloaded from the Apple App Store or the Google Play Store and then opened by the user a download is measured, on the basis of one download per each user account. App updates are not counted as a download. A new download is recorded if a user deletes and re-installs the app, or if the user installs the app on a new mobile telephone. Downloading does not necessarily equate with current app usage because, amongst other things, an app may be downloaded years earlier and then not used for some time, or it may be recently downloaded but rarely used.

157    Domain claims that the evidence shows that the Domain app was downloaded more times than the REA app, and at a higher monthly average, for the calendar year 2015, for the period July to December 2015, and for the period February 2015 to January 2016. It submits that its app was downloaded more times in the period July 2015 to January 2016 but, on my understanding of the evidence, that claim is inconsistent with the evidence of Mr McNabb, its Audience Insights Manager. It argues that this is a legitimate basis for a claim (if made) that the Domain app is the most widely or commonly used property app.

158    Against this, REA relies on unchallenged evidence which shows that:

(a)    as at October 2015 the REA app had been downloaded approximately 4.807 million times while the Domain app had only been downloaded approximately 4.2 million times;

(b)    as at January 2016 REA’s app had been downloaded approximately 5.088 million times. Domain did not proffer a total in this regard but it appears from the monthly download figures in evidence that the total downloads of its app fall short of REA’s total; and

(c)    in the five months preceding the Advertisements the REA app had been downloaded 476,826 times while the Domain app had been downloaded approximately 19,000 times less. In the two months preceding the Advertisements the REA app had been downloaded 197,995 times and the Domain app 154,312 times.

159    In my view, the evidence tends to show that, although the Domain app had gained significant ground, the REA app had been downloaded more times overall including in the months immediately preceding the Advertisements. If it were necessary to decide I would not be satisfied that the evidence regarding the number of downloads supports a claim that, at the relevant time, the Domain app was the most commonly or widely used property app.

160    Accordingly, if (contrary to my view) the Most Used Representation was conveyed by the Second, Third or Fifth Advertisements, that representation was false or misleading and Domain’s conduct was misleading or deceptive or likely to mislead or deceive.

The Independently Ranked Representation

161    To the extent that Domain sought to rely upon the All Versions Ratings to justify its claim that it has the number one property app in Australia, REA argues that claim is misleading because consumer ratings are not a reliable indicator as they are not independent, objective or reliable.

162    I have no difficulty in accepting that the consumer ratings on the Apple App Store and the Google Play Store cannot be accepted as reliable when, amongst other things:

(a)    anyone can rate an app, and the persons who rate them are self-selecting;

(b)    consumer ratings of apps are open to manipulation, both positive and negative. Each party hinted that the other side sought to manipulate its rating;

(c)    it cannot be said with precision what the star ratings represent. For example, one user may give a high rating based on the design of the app, whereas another user may give a low rating because the app crashed on the day he or she was using it. The only way to know the basis for the star ranking is if the user leaves a comment explaining the ranking; and

(d)    the ratings do not ask the consumer to compare the app to another similar app and any comparison of rankings must be imprecise. Amongst other things, one person’s five star rating may be equivalent to another person’s three star rating.

163    However, this does not mean it was misleading for Domain to rely upon such ratings. The All Versions Ratings of the Domain and REA apps are ratings made directly by a large numbers of consumers over a lengthy period of time. While I accept the possibility that the parties made (illegitimate) attempts to boost their own ratings, given the volume of consumer ratings any such attempts would only have an effect at the edges. I infer that the ratings are made largely independently of the parties and that they are objective in the sense that they are made by app users. As I have said, apps are ubiquitous in modern life and the ordinary or reasonable member of either target class would have downloaded and used them. He or she will be aware that they are rated by consumers, that such ratings are only a guide and that they carry little assurance of reliability. It is unrealistic to treat the ordinary or reasonable reader of the Advertisements as being so naive or unworldly that he or she would think that a No. 1 rating would not be advertised unless the rating was objective, independent and reliable in the way REA contends.

164    Having said this, if (contrary to my view) the Second, Third and Fifth Advertisements conveyed the Independently Ranked Representation, I could not be satisfied that the Domain app was ranked as the No.1 property app in Australia by an independent, objective and reliable body. It is uncontentious that it was not. I would conclude that the representation was false or misleading and Domain’s conduct was misleading or deceptive or likely to mislead or deceive.

Domain’s other contentions in relation to the No.1 Property App Claim

165    Domain further contends that there are other grounds upon which it could legitimately claim that it has the No.1 property app in Australia including superior All Versions Ratings of the Domain app compared to the REA app on the Apple App Store and the Google Play Store, as well as various industry awards received by Domain for its app.

166    I will deal with the evidence in relation to the All Versions Ratings of the competing apps when dealing with the second phase of the advertising campaign. It suffices to note that the evidence shows that for a significant period the Domain app has achieved higher All Versions Ratings than the REA app. However, if (contrary to my view) the No.1 Property App Claim would be understood as conveying a definite representation of superiority upon which the ordinary member of the target audience could reasonably rely, I would not see the consumer ratings as sufficient to legitimise that claim. I would conclude that the representation was false or misleading and constitutes misleading or deceptive conduct or conduct which was likely to mislead or deceive.

167    In relation to the industry awards, the evidence shows that in 2015 the Domain app received the following industry awards in the Australian App Design Awards:

(a)    the Gold Award for the Domain Home Price Guide: Best New Service or Application;

(b)    the Gold Award for Australia’s Best Property App: Best Mobile Marketing Campaign;

(c)    the Silver Award for Home and Living: Domain Apps and website products;

(d)    the Silver Award for Wearable Apps: Best Connected Device and Wearable Tech; and

(e)    the overall Design Champion Award.

168    Mr Ruiz sought to diminish the significance of these awards, but I found that aspect of his evidence unpersuasive. If it were necessary to decide I would be satisfied that the Domain app has some features or functionality which made it legitimate to claim, amongst other things, that it was well-designed and the best new app in 2015. However, it is not clear to me to which of the alleged representations this evidence was directed. It is difficult to see how the fact that the Domain app achieved these awards is of relevance to whether any of the Combined Representation, the Most Listings Representation, the Most Used Representation or the Independently Ranked Representation were conveyed, or if conveyed whether the publication of one or other of those representations constitutes misleading or deceptive conduct.

G.    The Second Phase: THE FOURTH, SIXTH AND SEVENTH ADVERTISEMENTS

169    The Fourth, Sixth and Seventh Advertisements were published as part of the second phase of the advertising campaign, and they contain the same central claim.

Publication of the Advertisements

The Fourth Advertisement

170    The Fourth Advertisement comprised a one-page advertisement which replaced the front page of the 21 February 2016 edition of The Sunday Age newspaper. An image of the Advertisement appears below.

171    The Advertisement was similar in design and layout to the first page of the Second and Third Advertisements. The front page had the usual header of the newspaper at the top of the page, below which a prominent heading in white font on a green background appeared, which said “Should you sell?” Below that a prominent subheading in white font appeared which featured the expression, “Download Australia’s highest-rated property app” (the Highest Rated Claim).

172    Similarly to the Second and Third Advertisements, below these headings an image of a mobile telephone appeared with the Domain app displayed on its screen, an invitation to download the app, the logos for the Apple App Store and the Google Play Store (at which the reader could download the Domain app), and the Domain logo.

173    As is apparent from the image above, an asterisk was appended to the subheading “Download Australia’s highest-rated property app”. The asterisk was larger and easier to see than the caret in the Second and Third Advertisements and in my view it would have been apparent to the ordinary or reasonable reader. The asterisk indicated the existence of a footnote at the bottom of the same page which said in small but readable font:

Rated by over 21,000 consumers, All Versions rating via Google Play Store and App Store, as at 18 February 2016. 9,816 consumers have rated the Domain Property App the highest rating possible of 5 stars, 5,825 Google Play Store consumers and 3,991 App Store consumers, as at 18 February 2016.

The Sixth Advertisement

174    The Sixth Advertisement comprised a series of billboards located in NSW, Victoria and Queensland from 21 February 2016, an image of which appears below:

175    The billboards featured the headline “Download Australia’s highest-rated property app” (i.e. the Highest Rated Claim) above an image of mobile telephone upon which the Domain app was displayed, together with the Domain logo.

176    An asterisk was visible at the end of the headline which related to a footnote at the base of the billboard. It stated in small font:

Rated by over 21,000 consumers, All Versions rating via Google Play Store and App Store, as at 18 Feb 2016.

The Seventh Advertisement

177    The Seventh Advertisement comprised an advertisement on Facebook commencing on 21 February 2016, an image of which appears below:

178    A prominent subheading in the advertisement stated “Find your next home with Australia’s highest rated property app. Download today” (i.e. a variation of the Highest Rated Claim). Below this appeared an image of an attractive property and the words “Download Australia’s best property app”.

179    The Facebook advertisement did not contain a footnote qualifying the Highest Rated Claim. Ms Cruickshank testified that the claim was based on the same consumer ratings as the Sixth Advertisement but the footnote was not included because Facebook advertisements have restricted formatting and a limited number of characters.

The alleged representations

180    Each of the Fourth, Sixth and Seventh Advertisements prominently claimed that the Domain app is “Australia’s highest rated property app”. REA contends that this claim conveyed one or more of the following three representations:

(a)    that the Domain app which is available for download has been rated more highly than any other Australian property app (the Highest Rated Representation);

(b)    that the Domain app which is available for download has been rated as “Australia’s highest-rated property app” by an independent, objective and reliable source (the Independently Rated Representation); and/or

(c)    that the Domain app which is available for download has been rated as “Australia’s highest-rated property app” by consumers (the Consumer Rated Representation).

Whether the alleged representations were conveyed by the Fourth, Sixth and Seventh Advertisements

The Highest Rated Representation and the Consumer Rated Representation

181    REA contends that on a natural reading the phrases “Download Australia’s highest-rated property app” and “Find your next home with Australia’s highest-rated property app. Download today” in conjunction with the Domain branding, conveyed the representation that the Domain app has been rated more highly than any other Australian property app. I have no difficulty in accepting that, nor in accepting that it was available to the ordinary or reasonable reader to understand those phrases as conveying the meaning that such ratings had been made by consumers. However, REA’s alleged representations go further than that. REA contends that those representations were specific (and limited to) the Domain app which was then available for download (i.e. the most recent update).

182    To understand the controversy regarding these alleged representations it is necessary to understand the system for consumer ratings of apps on the Apple App Store and the Google Play Store. Following the release of an app, from time to time app developers will release “updates” of the app to enhance its appearance and/or functionality. When a consumer downloads an app or an updated version of an app, or when a consumer is prompted to do so by the app, the consumer is provided the opportunity to assign a rating between one and five stars to signal the degree of the consumer’s satisfaction with the app. The Apple App and the Google Play Stores collate these ratings, accord the app an average star ranking out of five stars, and publish the average ratings on the App Stores.

183    Each update creates a new version of the app. The evidence shows that REA and Domain each released new versions of their apps in February 2016, had released new versions every few weeks going back to early 2014, and in one instance REA updated its app seven times in less than seven weeks. Only the most recently updated version is available for download from the App Stores.

184    The Apple App Store assigns two ratings to each app. The first rating relates to the current version of the app (the Current Version Rating). The second rating is the average of consumer ratings across the apps’ lifespan and encompasses all versions of the app (which I have called the All Versions Rating). On the Apple App Store the All Versions Rating and the Current Version Rating can both be viewed. On the Google Play Store only the All Versions Rating can be viewed.

185    REA argues that the Advertisements referred only to the current version of the app because the Advertisement exhorted the reader to download the app and only the current version was available for download. It also notes that the features of the app that the Advertisements promoted were features of the current version. It says that the representation as to the rating of the Domain app must relate to the current version because consumers will understand that the All Versions Rating is not a reliable indicator of consumer satisfaction, it being based on past versions of an app which have been superseded and which are no longer available for download.

186    I do not accept REA’s contentions. In my view the Fourth, Sixth and Seventh Advertisements did not convey the Highest Rated Representation or the Consumer Rated Representation in the terms alleged by REA.

187    In relation to the Fourth Advertisement I say this because, unlike the caret appended to the earlier Advertisements, the asterisk appended to the Highest Rated Claim was plainly visible. It was sufficiently prominent to indicate the existence of the footnote to the ordinary or reasonable reader, and the footnote was sufficiently prominent and clear in its terms. The footnote operated to qualify rather than undercut the primary message of the Advertisement. I consider the ordinary or reasonable member of each target class would have read the footnote and understood the qualifying information.

188    Having regard to the qualifying information, the Fourth Advertisement conveyed the representation that the Domain app has been rated more highly than any other Australian property app based on the All Versions Rating assigned to the app by consumers via the Apple App Store and the Google Play Store as at 18 February 2016. This representation was not specific (or limited) to the most recent update of the app.

189    Some different considerations apply to the Sixth Advertisement because the footnote to the billboards was unlikely to have been taken in by passing car drivers or pedestrians, and also to the Seventh Advertisement on Facebook because that Advertisement did not contain a footnote. In those Advertisements I consider the Highest Rated Claim conveyed a representation to the ordinary or reasonable reader that the Domain app was rated more highly than any other comparable property listing app by some rating process or system, probably by reference to consumer ratings.

190    That representation would not be understood as specific (or limited) to the latest update. As I have said, apps are ubiquitous in modern life and the ordinary or reasonable member of either target class would know that they carry consumer ratings and that well-used apps are regularly updated. He or she is likely to understand the Highest Rated Claim as referring to consumer ratings made over a reasonable period of time rather than just since the latest update. He or she would understand that a Current Version Rating would relate to only a relatively short period and would be inapplicable some short time later when the app was again updated.

191    I take the same view in relation to the Consumer Rated Representation. That representation would not be understood as specific (or limited) to the latest update. I do not accept that the ordinary or reasonable member of either target class would understand the All Versions Rating to be unreliable or obsolete, such that he or she would think that the Current Version Rating would be used. To the extent the ordinary reader would give the claim this much attention, he or she would know that because updates are regularly released the Current Versions Rating of an app will be based on a much smaller sample of consumers and be inherently less reliable than the All Versions Rating which is derived over a longer period of time from a significantly larger sample. The ordinary or reasonable reader may take a different approach if he or she was aware that there had been some wholesale change to the app as a result of the latest update, but that is not the evidence in the present case. The evidence tends to show that Domain made incremental changes to its app over a significant time period.

The Independently Rated Representation

192    REA contends that the Highest Rated Claim conveyed the representation that the Domain app then available for download (i.e. the most recently updated version) was rated more highly than any other Australian property app by an independent, objective and reliable source, particularly in relation to the Facebook advertisement (which did not include the footnote) and the billboard advertisement (where the footnote would not be taken in by passers-by). It argues that the prominence and the specific nature of that claim meant that an ordinary or reasonable reader was likely to assume that it could only legitimately be made with a firm evidentiary foundation, and that the rating must therefore have been carried out by an external source which is independent, objective and reliable.

193    I do not accept REA’s contentions. First, for the reasons already given, I consider that the ordinary or reasonable reader would not understand the claim that the Domain app is the highest rated property app in Australia to be specific and limited to the most recent update of the app.

194    Second, for similar reasons to those underpinning my conclusion that the Advertisements in the first phase of the advertising campaign did not convey the Independently Ranked Representation (see [129]-[130] above), I do not accept that the Advertisements in the second phase of the advertising campaign conveyed the Independently Rated Representation. I see it as quite unlikely that the ordinary or reasonable member of either target class would reach the view that there was some unidentified independent body which rated property apps against some unknown criteria. It is uncontroversial that no such body exists, and I am not satisfied that the ordinary reader would think that it did. Any person that has downloaded and used apps would know that the only ratings provided on the App Stores are consumer ratings.

Whether (if the representations were conveyed) they were false or misleading or constitute misleading or deceptive conduct

195    Having found that the Fourth, Sixth and Seventh Advertisements did not convey the alleged representations it is unnecessary to decide whether any such representations would be false or misleading or constitute misleading or deceptive conduct. However, given that the case was fully argued and having regard to the possibility of an appeal, it is again best that I briefly set out my views in this regard. It is also necessary to deal with REA’s contention that, even if the Highest Rated Representation and the Consumer Rated Representation are not limited to the current version of the Domain app, those representations were false or misleading.

The Highest Rated Representation and the Consumer Rated Representation

196    Both parties put on evidence as to the consumer ratings of the apps on the Apple App Store and the Google Play Store. REA relies on evidence regarding the Current Version Ratings of the competing apps which shows that:

(a)    as at 11 February 2016 the Current Version Rating of the REA app in the Apple App Store was 3.5 stars, compared to 3 stars for the Domain app; and

(b)    as at 21 February 2016 the Current Version Rating of both apps was 3.5 stars.

197    The evidence shows that at the time the Advertisements were published the REA app had a Current Version Rating which was either equal to or higher than the Current Version Rating achieved by the Domain app. Accordingly, if (contrary to my view) the Highest Rated Representation and the Consumer Rated Representation were specific to the most recent update of the Domain app then those representations were false or misleading and Domain’s conduct was misleading or deceptive or likely to mislead or deceive.

198    I do not however accept REA’s contention that the Highest Rated Representation and the Consumer Rated Representation were false or misleading if the representations were not specific to the latest update of the Domain app.

199    I say this, first, because the evidence in relation to the comparative All Versions Ratings of the competing apps shows that for a period of about 12 months prior to the publication of the Advertisements the Domain app had a superior All Versions Rating to the REA app in both the Apple App and Google Play Stores. As at 12 February 2016:

(a)    in the Apple App Store, the Domain app had an average All Versions Rating of 3.5 stars compared to 3 stars for the REA app. In the Google Play Store, the Domain app had an average All Versions Rating of 4.2 stars compared to 3.3 stars for the REA app;

(b)    in the Apple App Store the average All Versions Rating for the Domain app was based on 10,579 consumer ratings, whereas that rating for the REA app was based on only 2,720 consumer ratings. Similarly, in the Google Play Store the average All Versions Rating for the Domain app was based on 11,000 consumer ratings, whereas that rating for the REA app was based on only 6,832 consumer ratings. That is, the All Versions Ratings of the Domain app were based on many more consumer ratings and it maintained a higher rating than the REA app; and

(c)    in the Apple App Store, the Domain app had achieved the highest consumer rating on 4.5 times more occasions than the REA app (3,983 users compared to 897 users). In the Google Play Store, the Domain app had achieved the highest consumer rating on almost 2.5 times more occasions than the REA app (5,783 users compared to 2,319 users).

As at 23 February 2016 the Domain app continued to lead the REA app in app store ratings. Based on the All Versions Ratings it is plain that the Domain app had a clear ascendancy over the REA app for a significant period preceding the Advertisements.

200    Second, I do not accept REA’s contention that the Current Version Rating of the apps is superior to the All Versions Rating, such that Domain’s reliance on the All Versions Rating would be misleading. Because updates of both apps are released every few weeks the Current Version Rating will fluctuate significantly based on a small number of recent consumer ratings and it can be susceptible to manipulation. It will be inherently less reliable than the All Versions Rating which is derived over a longer period of time from a significantly larger sample of consumers. It is worth noting that the Google Play Store does not even publish a Current Version Rating. The ordinary or reasonable member of each target class would know that well-used apps are regularly updated.

201    Third, I do not accept that the billboard Advertisement (where the ordinary reader would not take in the footnote) and the Facebook advertisement (which did not include a footnote) were misleading because they did not distinguish between the All Versions and Current Versions Ratings and the reader could therefore be misled as to the basis of the representation. For the reasons already given I do not consider that the ordinary or reasonable member of either target class would understand the Highest Rated Representation or the Consumer Rated Representation to be based on the Current Version Rating.

202    Fourth, although the Domain app has been upgraded on many occasions it does not follow that the majority of reviews that form the basis for its All Versions Ratings are based on versions of the app which are so different from the most recent version that any comparison would be unfair. I accept Ms Cruickshank’s evidence that a significant proportion of the All Version Ratings were made in the two years prior to publication of the Advertisements, in which period many of the improvements to the Domain app were made. In considering the reliability of a consumer rating it is likely that the ordinary or reasonable member of each target class would see it as relevant to understand the number of consumer ratings made over some reasonable period of time.

203    Fifth, I do not accept REA’s contention that the representations were false or misleading because there are other property apps available in the app stores which have an All Versions Rating equivalent to and in some cases better than the Domain app. I accept Ms Cruickshank’s evidence that the other property apps to which REA refers are not comparable to the Domain app having regard to the services they provide, the markets in which they compete and the comparatively small number of ratings they have received. The relevant comparator for the Domain app (and no doubt the target of the claim to be the highest rated property app) was the REA app.

The Independently Rated Representation

204    The Domain app was not rated by an independent, objective and reliable external body or source. Accordingly, if (contrary to my view) the Fourth, Sixth or Seventh Advertisements conveyed the Independently Rated Representation, in publishing those Advertisements Domain made false or misleading representations and its conduct was misleading or deceptive or likely to mislead or deceive.

CONCLUSION

205    The parties are directed to confer in attempt to agree on draft minutes of orders to reflect these reasons, and to set out a timetable for dealing with the issues of relief and contempt of Court. Within 14 days the parties shall file agreed draft minutes of orders, or in the absence of agreement each party shall file draft minutes of the orders it seeks.

Associate:

Dated:    15 February 2017