Wednesday, July 17, 2013

Test Your TTAB Judge-Ability: Are Prepared Alcoholic Cocktails Related to Beer?

The PTO refused registration of the mark BATTLEFIELD BLUE 'RITA for "prepared alcoholic cocktail[s]" [BLUE disclaimed], finding the mark likely to cause confusion with the registered mark BATTLEFIELD for "beer." The Board deemed the marks to be "highly similar," since applicant "has essentially appropriated the registrant's entire arbitrary mark." But what about the goods? Are they sufficiently related for Section 2(d) purposes? In re Genghis Grill Franchise Concepts, LP, Serial No. 85372951 (July 2, 2013) [not precedential].

The Board noted that on many occasions different types of alcoholic beverages have been found by its reviewing court and by the Board to be sufficiently related in a likelihood of confusion analysis, because the beverages may share common uses, channels of trade, and consumers. For example: beer and ale related to tequila; malt liquor related to tequila; beer to wine; whiskey to gin and vodka; cognac to malt liquor, beer and ale; distilled spirits to wine; wine to whiskey; and Scotch whisky to rum.

Examining Attorney Linda E. Blohm submitted evidence that beer and alcoholic cocktails may be offered in the same channels of trade to the same customers: specifically, she provided retail store advertisements and restaurant menus where beer and alcoholic cocktails appear alongside each other. News articles, advertisements, and recipes demonstrated that beer may be used as an ingredient of alcoholic cocktails.

The Board disregarded applicant's arguments that its prepared cocktails will be served only in restaurants and that registrant's beer will be offered at registrant's brewery, pubs, and the like, since there are no such restrictions on channels of trade in the application and cited registration.

And so the Board affirmed the refusal.

Read comments and post your comment here.

TTABlog note: Remember, there is no per se rule that ....

Text Copyright John L. Welch 2013.


At 11:29 AM, Anonymous Mark L. Donahey said...

Ugh. This decision doesn't reflect the reality of the marketplace for the respective goods. There is no persuasive evidence suggesting that consumers are conditioned to expect beer and alcoholic cocktail goods to be offered under the same mark by the same source. For example, there are 545 live, use based registrations that identify "prepared alcoholic cocktail[s]" in Class 33, and there are 5609 that identify "beer[s]" in Class 32. Yet, only 23 of these live, use based registrations identify both "prepared alcoholic cocktail[s]" and "beer[s]". That is, 95.8% of alcoholic cocktail registrations don't include beer, and 99.6% of beer registrations don't include alcoholic cocktails.

Therefore, the consistency of the Board's decisions in these cases has much more to do with applicants' consistent failure to come forward with actual evidence challenging the relatedness of the goods than any genuine relationship between the goods themselves.

So what do you offer to support this conclusion? As outlined in the excellent article by Walsh and Hall (available from the TTABlog HERE), "Just as examining attorneys may use third-party registrations to show relatedness, 'applicant's may submit sets of third-party registrations to suggest the opposite, i.e., that the Office has registered the same mark to different parties for the goods at issue.' In re G.B.I. Tiles and Stone, Inc., 92 U.S.P.Q.2d 1366, 1369-70 (T.T.A.B. 2009)." In this case, the trademark database includes several examples of registrations for similar marks, in which the first was issued to one party for use in connection with "prepared alcoholic cocktail[s]" while the second was issued to a different party for use in connection with "beer[s]".

At 11:33 AM, Anonymous Mark L. Donahey said...

In support of my previous post, I provide eight such examples of coexisting marks for "prepared alcoholic cocktail[s]" on one hand, and "beer[s]" on the other. There are several others too.
8. 190 KNOCKOUT ["190" disclaimed] and KNOCKOUT BLOND ["BLOND" disclaimed].

This evidence suggests that the PTO often permits similar marks from different sources to coexist on the register for the relevant goods and the registrants themselves do not consider the relevant goods to be so closely related. In our experience, such evidence, if offered, often serves to rebut the examining attorney's prima facie case of relatedness, resulting in withdrawal of the refusal.

So, let's all start offering better evidence to challenge relatedness so that examining attorneys and the Board will have a better opportunity to issue common sense decisions based on the reality of the marketplace.

At 12:22 PM, Anonymous Freiburger said...

It bothers me when the Board uses the verb "appropriate" because it's a somewhat loaded term. Why not use a more neutral term like "contain," as in "the applicant's mark contains the registrant's entire arbitrary mark."

At 5:04 PM, Anonymous Anonymous said...

In the spirit of everyone "offering better evidence to challenge relatedness" -- Mr. Donahey please share with us your search methodology. Are you going through all search results for the "545 live, use based registrations that identify "prepared alcoholic cocktail[s]" in Class 33" and "5609 that identify "beer[s]" in Class 32"? and looking for similarities?

Seems very time consuming for these particular goods. Yet you seem to have done it fairly quickly for a comment in a blog post.


Are you using the TESS search functions on the USPTO website? or something else?

At 12:21 PM, Anonymous Mark L. Donahey said...

Ah, there's the rub. There is no efficient TESS search function for this. I wrote some custom software of my own to get this done.

The brute force method you ask about is hard but not impossible if at least one of the two categories of goods/services has relatively few registrations in it. In this case, you'd want to select terms from the smaller list of 545 cocktail registrations and look for matches in the longer list of beers. It's still very time consuming when dealing with small lists, and 545 registrations really isn't all that small. As the lists grow, the problem quickly becomes intractable by hand. Don't even think about a manual comparison of 50,000 t-shirt regs and 25,000 restaurant regs.

To be clear, I didn't intend my comment to come off as an indictment of applicants that don't introduce this kind of evidence. I probably appreciate the difficulty of finding it better than anyone. This kind of case is actually relatively rare, and many times the case may not be worth investing the resources necessary to conduct a rigorous search for these examples.

Rather, my concern has always been with the false impression these cases seem to promote in the trademark community. John is very good about reminding his readers that there is no per se rule that any two goods are related, but I fear that too many practitioners still walk away with the shorthand conclusion that goods X and Y are just always related. However, the fact is that a different record in these cases would still yield a different outcome. As long as we continue to keep that in mind, we can get better outcomes for clients who are willing to invest in developing the record, and we can avoid wasting the money of clients are aren't. But I digress.

At any rate, our docket at Pelton is large enough that we run into these cases a few times a year. A few years ago I began to get frustrated by the disproportionate ease of finding third-party registrations supporting relatedness compared to the intense challenge of finding evidence to rebut them. I started tinkering with software solutions on my own, and I eventually put together a system that was efficient enough to employ on behalf of our clients. It's not exactly quick, simple, or user friendly, and it doesn't entirely eliminate the need for subjective legal judgments. However, it speeds things along tremendously, and it allows us to do things that would otherwise be impossible with TESS or other conventional tools. If you need help on one of these comparisons, feel free to get in touch with us. I could probably help you out.

At 4:53 PM, Anonymous Anonymous said...

Actually, using TESS can get you the same results in a few minutes. Search "(032)[ic] & (live)[ld] & (beer or beers)[gs] & (1a)[cb] not (0)[rd]" (5627 records, close enough to what Mr. Donahey achieved), then search "(033)[ic] & (live)[ld] & ("prepared alcoholic cocktail")[gs] & (1a)[cb] not (0)[rd]" (545 results), and then search "(033 & 032)[ic] & (live)[ld] & ("prepared alcoholic cocktail" & beer)[gs] &(1a)[cb] not (0)[rd]" (24 results). I won't bother promoting my firm or my expertise -- I don't think this is an appropriate forum for that.

At 5:36 PM, Anonymous Anonymous said...

Mr. Donahey's examples are largely unconvincing given the different commercial impressions created by the marks. For example, 190 KNOCKOUT does not create the same impression as KNOCKOUT BLOND (a double entendre). If the eight referenced examples consisted primarily of marks like KNOCKOUT versus KNOCKOUT BEER, his third party regs would be more persuasive.

At 6:57 PM, Anonymous Mark L. Donahey said...

Anonymous@4:53: Yes, the TESS searches you propose will yield the full of all registrations for prepared alcoholic cocktail[s], the list for all beer[s], and the list of registrations that include both. I used something similar to get my initial totals as well. However, I believe that the previous anonymous commenter @5:04 was asking about the next step: Finding pairs of registrations for similar marks in which (i) each is owned by different parties, (ii) one identifies prepared alcoholic cocktail[s] but not beer[s], and (iii) the other identifies beer[s] but not prepared alcoholic cocktail[s]. If you know how to do this using TESS, I'd love to learn how.

Anonymous@5:36: I absolutely agree that the examples I provided could be more compelling, but I still think they are relatively persuasive under the circumstances.

First, the number of high quality pairs available to be found in the register is dictated by an interaction of two factors: (i) the degree of relatedness between the two categories of goods, and (ii) the number of registrations in each category to start with. Therefore, even if two categories of goods were totally unrelated, one couldn't necessarily count on finding any similar pairs if the number of registrations in one or both categories is relatively small. Here, there are only 545 registrations for "prepared alcoholic cocktails", so there just aren't that many opportunities for a collision in the first place. After all, there are millions of words in the English language alone, so there's lots of room for marks to miss each other. The list of pairs gets a lot longer and the matches get more compelling when we compare beer registrations with the much longer set of 16,000+ wine registrations. At any rate, I'm working on a model to quantify the degree to which the size of the lists affects the number of matches one should expect to find. This would help establish a yard stick for evaluating just how persuasive the eight examples I provided should be.

Second, even though many of my eight examples are not perfect, I hope you will admit that they are more persuasive than offering nothing at all. In most of these cases, applicants end up offering mere argument about factors that the Board simply cannot consider. In contrast, the Board has recognized that the eight paris have at least some persuasive value. They don't necessarily have to be overwhelmingly convincing to work. They just have to rebut the often limited case of relatedness presented by the examining attorney. Frankly, I suspect that in many cases the Board would prefer to find the relevant goods unrelated except that the applicant hasn't given them anything to hang their hats on. At minimum, this evidence starts to fill the void.


Post a Comment

<< Home