Peek & Cloppenburg KG, Düsseldorf, sought to register SAVANNAH for clothing in Switzerland. The Swiss IPO refused, citing the concern that the relevant public (general population) would understand SAVANNAH as referring to the town of Savannah, Georgia, USA, and that the mark was therefore misleading, as the goods do not necessarily originate from the US (the usual strange Swiss way of dealing with geographical indications as marks).

The Federal Administrative Court corrects. Although it thrashes a survey submitted by the applicant which purports to show that a majority of the population understands SAVANNAH primarily as a form of vegetation because the sample size of 504 is too small ( a 1,000 is required) and the question format not exactly to the liking of the court, it comes to the very same conclusion as the demoscopic survey based on its own common sense ("Lebenserfahrung"): SAVANNAH will be primarily understood as reference to a form of vegetation by the Swiss public, and not as reference to Savannah, Georgia. Hence, the mark may be protected, as SAVANNAH is not descriptive for the goods claimed.

Don't get me started on why it is wrong to think it is better to base a decision on one's "common sense" than on a (pretty good, btw) survey. "Common sense" or "Lebenserfahrung" is about the most overrated concept in trade mark law. No, make that, "in law". No, better, "in life". And claiming that a sample size of 500 leads to a drastically reduced weight of evidence compared to a sample size of 1,000 shows a lack of understanding of what sample sizes, and the resulting confidence intervals, means, but this is too complicated a topic for a blog post. Should the day ever have more than 24 hours, I shall write the article that I have been intending to write on this for a long time.

German summary of the decision of 18 July 2012 here (with link to full text). The decision could be appealed to the Supreme Court by the IPO, but as it turns on factual issues, I'd be surprised if it had been.