WEDNESDAY, 29 FEBRUARY 2012
Switzerland: Oktoberfest-Bier and good survey practices
The Verein Münchner Brauereien eV is a trade association comprising the Munich breweries Augustinerbräu, Hofbräuhaus, Spaten, Löwenbräu, Hacker-Pschorr and Paulaner. Its members are the only entities allowed to sell beer on the Wiesn during the Oktoberfest (nice business if you can get it). The assocation owns the German trade mark OKTOBERFEST-BIER for, yes indeed, beer, since 1982, and the eponimous Community Trade Mark since 2000 (registered based on acquired distinctiveness, no. 220772).
In 2008, the association tried to extend the protection of its mark to Switzerland as a collective mark, using the Madrid System. The Swiss IPO refused, noting that the Swiss public would not perceive OKTOBERFEST-BIER as an indication of source, but rather as a type of beer. The applicant tried to prove acquired distinctiveness with a demoscopic survey, and failed. On appeal, the Federal Administrative Court upheld the IPO's refusal to allow the registration of the mark which lacked distinctiveness. The survey evidence was considered insufficient because of methodological deficiencies. The case is interesting for practitioners because the court carefully addresses some issues with demoscopic surveys.
Some takeaway points:
in the case of "omnibus surveys" (i.e. several topics are combined in one interview), the questions immediately preceding the relevant question should also be reported, as they could influence the answers to the relevant questions;
a sample size of 1,000 interviews (actually answered questions) is generally required;
question "Does the sign OKTOBERFEST-BIER give an indication (einen
Hinweis geben) to one or several or no company (Unternehmen)?"cannot be answered with "yes" or "no", as there are actually three alternatives given;
if a product is marketed towards the general population, it is not permissible to use filter questions. In this case, the survey asked "Do you regularly consume beer?", and those (23%) of the interviewed that answered that they never drink beer where excluded from the subsequent analysis. The court insists that they must be counted towards the total, i.e., if you need to show that more than 50% of the general population associates your mark with one source, those persons not consuming the type of product your mark is seeking protection for must nonetheless be included when calculating the fraction.
One final rant: the court once again refused to tell us what the percentage of the relevant consumers is that needs to associate the mark with one source to show acquired distinctiveness. All it tells us is that it is a "substantial part", and the rest depends on the circumstances of the case. This is really unsatisfying, as it makes it entirely unpredictable how they will decide even when a demoscopic survey shows that more than 50% of the relevant consumers associate the mark with one source. Is it really that hard to just say that > 50% is enough??
In this case, the survey evidence could not prove that a substantial part of the general population in Switzerland associated OKTOBERFEST-BIER with a single source.
(German) summary of decision of 17 February 2012 (with link to full text). The decision could be appeald to the Supreme Court, but I would be surprised if it would (because the Supreme Court is bound by the facts as established by the lower court, and in this case, the decision turns on the facts).
Posted by: Mark Schweizer @ 10.48