It has now been a decade since the Supreme Court of Canada’s decision in Masterpiece v. Alavida 2011 SCC. In Masterpiece, Justice Rothstein had some harsh words regarding the use of expert evidence in trademark cases. At the time, many commentators predicted that the case would signal the death of surveys and other expert evidence in trademark cases. A review of the cases that have been decided since Masterpiece reveals that, although Canadian courts are giving closer scrutiny to trademark surveys, they are continuing to admit well-designed and well-conducted surveys. At the same time, however, courts have also used the Masterpiece decision as justification for rejecting surveys and other expert evidence where there are flaws in the evidence or the trier of fact concludes that they are well-positioned to rule on the issue of confusion without expert assistance.
The Masterpiece Decision
Masterpiece was a relatively straightforward trademark case and so a detailed review of the facts is not particularly instructive. What is instructive is the portion of Justice Rothstein’s decision addressing the issue of when courts should (and should not) take expert evidence into account in trademark cases.
Criteria for Acceptance of Expert Evidence
Justice Rothstein began his analysis by noting that tendering expert evidence in trademark cases is no different than tendering expert evidence in other cases. The following four requirements must be met before expert evidence can be accepted:
- necessity in assisting the trier of fact;
- the absence of any exclusionary rule; and
- a properly qualified expert.
In the context of trademark surveys, the first criterion has been interpreted to require that survey evidence be both reliable (i.e., the survey can be expected to produce consistent results if it is administered on multiple occasions to different individuals) and valid (i.e., the right questions are put to the right pool of respondents in the right way and the right circumstances).
The second criterion has been interpreted as prohibiting the testimony of an expert if their evidence is not likely to be outside the experience and knowledge of a judge. Justice Rothstein commented that the criterion of “necessity” is not met when an expert’s evidence would merely be “helpful” to the trier of fact.
Justice Rothstein then opined that, given the expensive nature of litigation, courts must fulfil their gatekeeper role to ensure that unnecessary, irrelevant, and potentially distracting expert and survey evidence is not allowed to extend and complicate court proceedings.
Prior to Masterpiece, it was common for a judge to admit survey evidence and then let any objections regarding the survey go to the weight to be given to such evidence. Justice Rothstein commented, however, that if a trial judge concludes that proposed expert evidence is unnecessary, irrelevant, or will distract from the issues to be decided, they should disallow such evidence from being introduced. Justice Rothstein further suggested that proposed expert and survey evidence should be a matter for consideration at the case management stage of proceedings so that the costs associated with fighting over the evidence at trial can be avoided.
Justice Rothstein stated as follows regarding the criterion of necessity:
In a case such as this, where the “casual consumer” is not expected to be particularly skilled or knowledgeable, and there is a resemblance between the marks, expert evidence which simply assesses that resemblance will not generally be necessary. And it will be positively unhelpful if the expert engages in an analysis that distracts from the hypothetical question of likelihood of confusion at the centre of the analysis.
Justice Rothstein next reviewed a number of cases that considered the role of expert evidence in trademark cases and commented as follows:
- where the market is specialized, evidence about the special knowledge or sophistication of the target consumers may be essential to determining when confusion would be likely to arise;
- however, where goods are sold to the general public for ordinary use, it is difficult to see what is gained from the evidence of an expert as to his or her own opinion since judges are entitled to given effect to their own opinions as to the likelihood of deception or confusion;
- judges should use their own common sense, excluding influences of their own idiosyncratic knowledge or temperament to determine whether the casual consumer would be likely to be confused; and
- surveys have the potential to provide empirical evidence which demonstrates consumer reactions in the marketplace and so can meet the necessity criterion for the admission of expert testimony.
Simulation of an Imperfect Recollection
Justice Rothstein noted that the facts in Masterpiece raised a particular problem for the admission of survey evidence. In the test for confusion under the Trademarks Act, the likelihood of confusion is assessed in the mind of the average consumer, having an imperfect recollection of party A’s trademark, who then encounters the trademark of party B. In such circumstances, there is a likelihood of confusion if the consumer would mistakenly conclude that the goods or services associated with those trademarks are manufactured, sold, leased, hired or performed by the same person.
In assessing the likelihood of confusion, subsection 6(5) of the Trademarks Act provides that a court or the Registrar of Trademarks shall have regard to all the surrounding circumstances, including among others the extent to which the trademarks have become known.
In a scenario where the plaintiff has extensively used its trademark in Canada, a confusion survey conducted among a sufficiently large number of people will constitute a good proxy for the level of recognition of the plaintiff’s trademark among the relevant consuming public in Canada. Similarly, the results of a properly conducted survey should constitute a good proxy for the “imperfect recollection” that the average consumer would have of the plaintiff’s trademark when they encounter the defendant’s trademark.
A problem arises, however, in the scenario where the plaintiff has not yet used its trademark in Canada. A trademark owner is not required to have used their trademark in Canada in order to obtain a Canadian trademark registration. Such a person is thereafter entitled to enforce their registration against a third party who uses a confusing trademark. However, if one were to conduct a survey of, say, 500 Canadians the results may show that none of those people is familiar with the plaintiff’s trademark since the mark has not been used in Canada. In such circumstances, the survey would underestimate the true level of confusion that would occur if consumers were to encounter the defendant’s trademark.
In Masterpiece, Masterpiece Inc. had not yet established a presence in Canada and so there were no casual or average consumers with an “imperfect recollection” to survey. Masterpiece’s survey expert therefore used a series of questions that attempted to establish a proxy for an imperfect recollection of its trademark and then tested how customers would react when exposed to the second mark. Justice Rothstein was dismissive of this type of survey, commenting as follows:
This is not asking questions “in the right way, in the right circumstances” to elicit evidence of how those with an imperfect recollection of Masterpiece Inc.’s marks would react to Alavida’s proposed mark. For a survey to be valid, it seems elementary that there must be some consumers who could have an imperfect recollection of the first mark. Simulating an “imperfect recollection” through a series of lead-up questions to consumers will rarely be seen as reliable and valid.
While I would not absolutely foreclose the possibility that a party may devise a valid survey in a case where a trade-mark user has not established a sufficient presence in the marketplace for consumers to have formed an imperfect recollection of its trade-mark, I would venture that it is highly unlikely that such a survey would meet the requirements of reliability and validity.
Application of Masterpiece Decision
In the decade since Masterpiece, Canadian courts and the Trademarks Opposition Board (the “TMOB”) have had many occasions to apply the Supreme Court’s guidance when deciding on the admissibility of expert evidence in trademark cases and the weight to be given to such evidence.
Valid and Reliable Surveys Still Welcome
By and large, the courts and the TMOB have been very mindful of the Supreme Court’s cautions in Masterpiece and often reference the decision at the outset of any discussion on the admissibility and weight of a survey or other expert evidence. Although the courts have not always applied the same level of strictness to the expert evidence before them, they have generally continued to be receptive to well-designed surveys that meet the criteria of validity and reliability.
For example, in Research in Motion Limited v. LG Electronics 2012 TMOB 253, the TMOB cited the Supreme Court’s commentary in Masterpiece regarding expert evidence and surveys and noted that the main difficulty with surveys is often their reliability and validity. The TMOB then went on to hold that the confusion survey before the Board was both valid and reliable and so was entitled to be given full weight.
Similarly, the court in Cheah v. McDonald’s Corporation 2013 FC 774 acknowledged the caution in Masterpiece that “survey evidence should be used with caution and not supplant the role of the judge”. It then went on to hold that there was no reason to fault the survey evidence in the case before it.
In Diageo Canada v. Heaven Hill Distilleries 2017 FC 571, the defendant brought a pre-trial motion to exclude the plaintiff’s confusion survey. The trial judge began his analysis of the defendant’s objections by citing not only the caution in Masterpiece against surveys but also the Supreme Court’s comment that surveys have the potential to provide empirical evidence which demonstrates consumer reactions in the marketplace. The judge then went on to hold that the survey before him was, despite the defendant’s objections, both reliable and valid and thus admissible.
However, there have also been many other cases where the courts and the TMOB have either refused to admit survey evidence or have given it reduced weight. The below discussion summarizes how survey evidence has fared, both good and bad, in trademark cases post-Masterpiece.
A Properly Qualified Expert
As noted above, one of the four criteria for the acceptance of expert evidence is that it must be tendered by a properly qualified expert.
In Wenger S.A. v. Travel Way Group 2016 FC 347, the court gave no weight to a survey because, among other things, the results were introduced through an affidavit sworn by a person who neither designed the survey nor conducted the interviews of the survey respondents.
In Credit Union Central of Canada v. CCUEFINANCE Consulting 2016 TMOB 175, one of the applicant’s affiants submitted copies of 18 completed survey questionnaires regarding the relationship between the opponent, applicant and their respective marks. The opponent challenged the admissibility of the survey on the grounds that it was not designed or conducted by a qualified expert but was instead designed by a friend of the affiant who did not have any particular qualification for conducting survey or opinion polling. Additionally, the survey was carried out by the applicant’s employees. Not surprisingly, the TMOB gave no weight to the survey although it only commented on the survey not meeting standards of relevance and reliability rather than also commenting on the fact that it was not designed and administered by a properly qualified expert.
In Arterra Wines Canada v. Diageo North America 2020 FC 508, the applicant relied on evidence of a Master Sommelier regarding, among other things, whether the public would be confused between the trademarks NAKED GRAPE and THE NAKED TURTLE when used in association with alcoholic beverages. The TMOB found that the affiant was not qualified to render an opinion as to confusion arising from the trademarks at issue or with respect to the inherent distinctiveness of the parties’ trademarks. Although the affiant was no doubt an expert on certain aspects of wine, he was not an expert in human behaviour and had not conducted surveys regarding what the average consumer of wines and spirits would understand regarding the meaning of the term “naked”. On appeal, a judge of the Federal Court agreed with the TMOB’s ruling, holding that the ruling was in alignment with the Supreme Court’s guidance in Masterpiece that the trier of fact “should use their own common sense, excluding influences of their ‘own idiosyncratic knowledge or temperament’ to determine whether the casual consumer would be likely to be confused.”
In Sobeys Capital Incorporated v. EDENRED, S.A., 2012 TMOB 87, the opponent led evidence that, as part of its commercial activities, it had commissioned a survey that showed that the opponent’s mark COMPLIMENTS was recognized by approximately 80% of the Canadian population. The opponent’s affiant, however, did not provide a copy of the survey as part of her evidence, taking the position that it contained confidential information. Not surprisingly, the TMOB declined to give any weight to this survey given that it was not filed in evidence by a qualified expert and there was also no way for the hearing officer to determine whether the survey was properly designed and conducted in an impartial manner.
Asking the Right Question
In Imperial Tobacco Canada Limited v. Philip Morris Brands, 2014 TMOB 223, Imperial Tobacco oppose Philip Morris’ application to register its so-called Roof design for cigarettes. The opponent submitted survey evidence that it had relied on in prior litigation between the parties. In deciding to give the surveys little to no weight, the TMOB commented that the surveys were not designed to test consumer reactions to the marks in the present opposition but rather to a no-name brand product made of a particular combination of various elements.
In another decision involving the same parties, Imperial Tobacco Canada Limited v. Philip Morris Brands, 2018 FC 503, a Federal Court judge was required to rule on survey evidence that had been submitted on appeals from several oppositions that had been decided by the TMOB. Not surprisingly, the survey evidence was held to be relevant to some marks and goods but not relevant to other marks and goods. As such, the survey evidence was admitted in some of the cases but rejected in other cases.
An important consideration in the design of a trademark survey is the manner in which the survey can measure a state of affairs (such as the reputation of a trademark or the likelihood of confusion) as of a particular date. It is not at all unusual in a trademark case for a survey to be conducted several years after the relevant date. Although it is often possible to design a survey so that it measures the state of affairs as of an earlier date (such as by asking respondents when they first become aware of a particular trademark), Canadian courts in several cases have had to consider whether survey evidence can be admitted when it measures reputation, confusion, etc. at a different date than the relevant date.
In Vancouver Community College v. Vancouver Career College, 2017 BCCA 41, the Court of Appeal considered the issue that, although the relevant date for assessing the likelihood of confusion was 2009, the appellant had tendered evidence of surveys conducted by it in 2005 and 2006 that addressed questions of name awareness. The court held that the surveys nevertheless provided some evidence of goodwill as of 2009 and, as such, should have been considered by the trial judge.
By way of contrast, in Ferring v. Apotex, 2013 TMOB 225, the TMOB Board gave no weight to a confusion survey that was conducted 5 years before the relevant date (the date of filing of the application that was being opposed).
As noted above, the Supreme Court of Canada in Masterpiece expressed skepticism that, in the scenario where a party had not used its trademark in Canada, it would be possible to design a survey that was based on a series of questions to establish a proxy for an imperfect recollection and only thereafter test how customers would react when exposed to the other party’s mark.
Subsequent to Masterpiece, the courts have had occasion to consider the admissibility and weight to be given to surveys that have attempted to simulate an imperfect recollection.
In Tokai v. Kingsford Products, 2021 FC 782, a Federal Court judge, in considering the admissibility of a survey noted that there were no issues with respect to the expert’s qualifications or the impartial and independent conduct of the survey. However, the judge went on to find that the survey did not do an adequate job of simulating a consumer having an imperfect recollection of the senior user’s mark since, among other things, the manner in which a survey participant was shown a brand name was not reflective of the manner in which the mark would be encountered in the marketplace.
Non-survey Expert Evidence
Cathay Pacific Airways v. Air Miles International, 2016 FC 1125 is a good example of the treatment given to non-survey expert evidence. In this case, evidence was tendered from a linguistics expert as to the degree of resemblance between the trademarks AIR MILES and ASIA MILES in appearance, sound and ideas suggested. The TMOB held that expert evidence which simply assesses resemblance will not generally be necessary.
Show Your Work
Canadian courts have consistently held that trademark survey evidence is only admissible if the party seeking to introduce the survey discloses all relevant documents with respect to the survey questionnaire and the conduct of the survey.
For example, in Clorox Company of Canada, Ltd. v. Chloretec s.e.c., 2018 FC 408 (CanLII) appeal dismissed 2020 FCA 76, a witness provided some charts that purported to show the results of a survey conducted on consumer awareness of certain “Javex” trademarks. These charts showed the proportion of persons who say they recognize certain trademarks of cleaning products. However, no information was provided on the survey’s methodology or on the identity of the firm that conducted it. In dismissing this evidence, the court commented as follows:
In this case, Ms. Hilt does not provide any information as to the methodology that was used to conduct these surveys. It is therefore impossible to assess their reliability and relevance to the questions that are at issue here. In fact, such evidence should have been submitted by means of the affidavit of an expert who conducted the survey and who provides sufficient explanations in relation to its methodology. Therefore, I will not consider these two surveys.
Let it Go to Weight, Not Admissibility
There have been many decisions since Masterpiece where, despite arguments from a party that the opposite party’s survey should be ruled inadmissible, the trier of fact instead admits the survey but gives it reduced weight due to the attacks on the survey.
For example, in Air Miles International Trading BV v. Cathay Pacific Airways Limited, 2012 TMOB 80, the applicant argued that the opponent’s surveys were inadmissible since the opponent’s experts did not provide sufficient background information concerning the conduct of the surveys. The Opposition Board held that the applicant’s arguments had some merit but ruled that those concerns went to the weight of the survey rather than to its admissibility.
Similarly, in Loblaws Inc. v. Columbia Insurance Company, 2019 FC 961, the judge ruled that some of the parties’ objections to each other’s expert evidence had merit but, for the most part, let the objections go the weight of such evidence rather than its admissibility.
In view of the above, it is fair to say that the Masterpiece decision has heightened awareness of the requirements for admissible expert evidence in trademark cases and has led to an improvement in the quality of evidence being tendered to the courts and the TMOB. It is also fair to say that a valid and reliable confusion survey will continue to often be the best available evidence on the issue.