Actual Dilution or Likelihood of Dilution? The Evolution of Trademark Law
an interview with John Crittenden of Cooley Godward Kronish
Heather Bromfield | University of California, Davis, School of Law
Jessica Richardson | University of California, Davis, School of Law

Posted Friday, May 2, 2008
8 U.C. Davis Bus. L.J. (2008)

1)     You were the lead litigator for eBay in the recently decided case, Perfumebay.com Inc. v. eBay Inc. (PerfumeBay). [1]Could you tell us a little more about this case and how the holding impacted eBay's senior trademark rights in the “eBay” name?[2]

Over the years, we have seen many e-commerce companies imitate the eBay trademark with names like “AutoBay,” “RentBay,” “JewelryBay,” “CoinBay,” and so forth, taking the generic name for what they sell and adding the “Bay” suffix of eBay. Some even mimic the colors of the eBay logo or the appearance of the eBay website. PerfumeBay's founder started by selling perfume online through eBay; four months later she decided to call her own website PerfumeBay.

     eBay takes action against these “generic plus Bay” names because they can confuse consumers into thinking that these sites are affiliated with eBay. And even if they don't confuse consumers, the proliferation of these names could dilute the distinctiveness of the famous eBay mark and perhaps ultimately make “Bay” a synonym for online marketplace.

     In the PerfumeBay case we brought claims for both trademark infringement (which is based on confusion) and trademark dilution. The part of the Ninth Circuit's decision on dilution is more interesting from a legal standpoint, so I'll discuss that first.

     After a bench trial, the district judge found that the eBay mark is famous, a key element of dilution. However, he did not find the PerfumeBay and eBay marks “substantially identical,” as case law required, and denied the dilution claim. We appealed that ruling.

     The Ninth Circuit reversed, holding that PerfumeBay was similar enough to eBay to dilute the mark. The panel held that the district court erred by looking at the marks' similarity in a vacuum. Instead, courts must consider similarity in the context of other factors, like the fame of the mark and how distinctive it is. Thus, where you have an extremely well-known and highly distinctive mark like eBay, less similarity is needed for a mark to dilute it. These factors are now codified in the new Federal Trademark Dilution Revision Act of 2006, which was enacted after our trial. On the infringement claim, the district judge's decision was split. He found a likelihood of confusion when PerfumeBay was conjoined as one word, but not if a space separates the two words. Both sides appealed.

     Unlike the purely legal issue of the dilution standard, the question here was whether the district court's fact finding was correct. Findings of fact are much harder to overturn than legal conclusions. The appellate court has the same access to the law as the district court does, so it can interpret the law de novo. But a finding of fact is based on hearing the evidence at trial, listening to the witnesses and experts, looking at the documents, and weighing all the facts. Because the appellate court didn't observe the trial itself, on factual matters it defers to the district judge who did. Only if it finds clear error will it upset a trial judge's findings of fact.

     We pointed out that many cases that hold that minor changes in spelling, spacing, or punctuation aren't enough to render a mark not infringing. Although the Ninth Circuit called our argument “persuasive,” it did not find the district judge's factual finding clearly erroneous, and upheld the ruling. Of course, the infringement ruling is confined to the facts of that particular case, and another court could find a “non-conjoined” generic plus Bay name to infringe the eBay mark.

2)     Is it fair to say that it is important to protect your mark in the very first instance of misuse? If you do not protect the mark in the first instance, are you out of luck in the future?

     It is important to protect your mark. I will not say that if you do not act in the first instance of infringement you are out of luck. However, if you do not show a good pattern of protecting your mark you can lose your rights in it. It ultimately depends on how the consumer perceives the trademark. For example, “Aspirin” is a registered trademark of Bayer in Canada, but here in the United States it is a generic term that anybody who makes that product can use. Originally, it was a trademark here, too, but other companies started calling their versions of the product “aspirin” too. People came to think of “aspirin” as a generic name rather than a brand of one company, and that's how the trademark rights were lost. The companies that make Xerox copiers, Kleenex tissues, Band-Aid bandages, and Rollerblades inline skates have been vigilant to prevent this from happening to their brands.

3)     In AMF Inc. v. Sleekcraft Boats, (Sleekcraft),[3] the Ninth Circuit Court identified eight factors relevant to the determination of the likelihood of trademark confusion: 1) the strength of the mark, 2) the proximity of the goods, 3) the similarity of the marks 4) the evidence of actual confusion, 5) the marketing channels used, 6) the type of goods and degree of care likely to be exercised by the purchaser, 7) defendant's intent in selecting the mark, and 8) the likelihood of expansion of the product lines. In consideration of your experience in cases such as PerfumeBay, what factors would you like to see added to or removed from to the Sleekcraft test to better resolve trademark issues and why?

     I believe the test is fine as is. Some circuits articulate it a little differently, but all circuits basically follow the same test. It is not a question of adding up the factors. For example, if all seven factors are in your favor, that does not necessarily mean there is infringement. Courts consider the factors flexibly. For example, if you have a relatively weak mark but you have almost identical products, confusion is more likely than if you have a weak mark and unrelated products. It is a nonexclusive list so other facts can be considered, although they seldom are. The factors simply provide a framework for courts to analyze claims just as the 2006 federal trademark dilution statute provides a nonexclusive list of things to consider when determining dilution. It is important to recognize these factors are interrelated, subjective, and not to be applied mechanically. I think that all the factors all work together well and it is generally understood, which makes for a good test.

4)     In Ringling Bros. Barnum & Bailey Combined Shows, Inc. v. Utah Division of Travel Development (Ringling Bros.),[4] the court created an additional economic hurdle for trademark owners to overcome before dilution can be found. Specifically, the court stated that there had to be an “actual economic harm to the famous mark's economic value by lessening its former selling power as an advertising agent for its goods or services” for dilution to have occurred.[5] There were two significant developments following Ringling Bros.: 1) Moseley v. V Secret Catalogue [6] and 2) the Trademark Dilution Revision Act of 2006.[7] In 2003, Moseley overruled Ringling Bros. economic harm requirement and affirmed the need for actual dilution. In 2006, the TDRA reversed Moseley's holding of required actual dilution because such a requirement proved too difficult to establish. Which approach do you prefer and why?

     The new dilution statute fixes the problem of the earlier law's “actual dilution” requirement. Proving that the mark has already been diluted is difficult and it almost suggests that a company must wait until its brand has already been damaged before it can even bring a claim. That was how some of the courts were interpreting the previous law. Conversely, other courts thought Congress did not expect companies to wait until their valuable brands were damaged before they could bring any kind of law suit. That is why some courts such as the Ninth Circuit applied a likelihood of dilution standard. After the Supreme Court in Moseley interpreted the language of the statute to mean actual dilution and not a likelihood of dilution, brand owners struggled with how to prove dilution.

     The Trademark Dilution Revision Act resolved this by making the standard likelihood of dilution. As I mentioned, it also provided a series of factors to consider in determining whether a mark is diluted, much like the likelihood of confusion factors in Sleekcraft.

5)     One of the purposes of the Lanham Act of 1946 [8] is to protect the consumer. In our society we look to protect the trademark owner from dilution as well as the consumer. Once the FTDA was passed in 1995, the number of law suits nearly doubled over the subsequent five years. As an expert in Alternative Dispute Resolution (ADR) do you think that there should be legislative or contractual developments to promote arbitration instead of costly and time consuming litigation? Or, do you think trademark dilution cases are best settled in the courtroom?

     If parties cannot resolve their case, they should have their day in court. That is a right the Constitution gives us, provided the jurisdictional requirements are met. However, I favor ADR. There are many barriers to settlement occur, such as advocate bias, unrealistic expectations, or when parties focus on positions rather than their needs or interests. A good mediator can help the parties find common ground and accommodate the interests of both sides. Because many of court systems already promote arbitration or ADR, there may not be a need for a legislative mandate.

     In the Northern District of California, for example, the court adopted a multi-option ADR program, one of the first of its kind in the nation. As part of the case management process, the court requires parties to talk and exchange information early and then participate in a mandatory non-binding ADR program. The parties do not have to accept the results of the ADR process–they still have their right to their day in court,–but they at least are required to go through the process. They can choose one of several types of ADR: arbitration, settlement conference, early neutral evaluation, or mediation.

     In courts like the Northern District of California, where there is strong support for ADR, many cases get settled early in the process. In other jurisdictions where ADR is not mandatory, lawyers may be less likely to propose ADR, fearing that may be considered a sign of weakness. Congress conceivably could direct all courts to have an ADR program like ours in the Northern District of California, and that could help reduce the number of cases that go to trial. However, every court is different. Smaller courts with few judges may not have the need or the resources to develop the infrastructure for this kind of program. While I do not necessarily see programs like the Northern District's being mandated on a national scale anytime soon, the trend is certainly in favor of ADR.


[1] Perfumebay.com Inc. v. eBay Inc., 506 F.3d 1165 (2007).

[2] This case asked the question of whether the PerfumeBay mark infringed (created a likelihood of confusion with) or diluted (diminished the distinctiveness of) the eBay mark.  The district court held that some versions infringed but that none diluted the eBay mark. The Ninth Circuit affirmed the ruling on infringement but also held that there was trademark dilution.

[3] AMF Inc. v. Sleekcraft Boats 599 F.2d 341, 348-49 (1979). (Comm. Print 1978).

[4] Ringling Bros. Barnum & Bailey Combined Shows, Inc. v. Utah Division of Travel Development, 170 F. 3d 449 (1999).

[5] Ringling Bros. Barnum & Bailey Combined Shows, Inc. v. Utah Division of Travel Development, 170 F. 3d 449, 461 (1999).

[6] Moseley v. V Secret Catalogue, 537 U.S. 418 (2003).

[7] 15 U.S.C.A. § 1125 (2006).

[8] 15 U.S.C. § 1051 (1946).

Heather Bromfield and Jessica Richardson are students at UC Davis, School of Law, and are Staff Editors on the Business Law Journal.