Notes and Questions on Playboy v. Netscape and Brookfield
Back to Top
1. The universal standard for trademark infringement (or for unfair
competition in the nature of trademark infringement) is likelihood of
confusion. See Jay Dratler, Jr., Intellectual Property Law: Commercial,
Creative and Industrial Property § 10.01 (Law Journal Press 1991,
updated semiannually). But confusion as to what? Over the
years the courts have expanded and categorized the types of confusion
that the trademark-infringement standard recognizes.
2. The oldest and simplest type of confusion is confusion as to
the source or origin of goods or services, i.e., as to who produces or
offers them. For example, in the paradigmatic case of "passing off,"
the defendant causes confusion by making customers think that its goods
or services come from someone else. See William R. Warner & Co.
v. Eli Lilly & Co., 265 U.S. 526, 527-528, 530, 44 S.Ct. 615, 68 L.Ed.
1161 (1924) (drug maker induced druggists to pass off its combination
of chocolate and quinine as plaintiff's unpatented mixture by selling
it wholesale at lower price and suggesting that druggists sell it as plaintiff's,
for higher price, and pocket the difference). This type of confusion
might be called "point-of-sale" confusion because the consumer buys a
product or service thinking it is produced or sold by someone other than
the actual producer or seller.
3. A second type of confusion is also easy to understand: confusion
as to sponsorship, affiliation or approval. If a well-known mark
is used for goods or services not produced by the trademark owner, the
purchaser may think that the trademark owner, although not actually producing
or selling the goods or services, sponsors or approves of them or is affiliated
with the actual producer. This type of confusion occurs commonly
in modern commerce because mergers, acquisitions, affiliated groups of
companies, and trademark licensing are so common. As a result, even
a well informed consumer might be led to think that there is some such
relationship between the offeror of a good or service and the trademark
owner, even if there is none. See Boston Professional Hockey Association
v. Dallas Cap & Emblem Manufacturing Co., 510 F.2d 1004, 1012 (5th Cir.),
cert. denied, 423 U.S. 868 (1975) (unauthorized use of National
Hockey League teams' insignia on cloth emblems created likelhood of confusion
as to source or sponsorship). Can you articulate precisely how this
sort of confusion harms consumers and mark owners? what trademark
policies it implicates?
4. Once one moves beyond confusion as to source, sponsorship, affiliation
or approval, however, the sorts of confusion that the law recognizes require
a little more explanation. One additional variety of confusion,
called "post-sale" confusion, depends on prospective purchasers
of similar goods bearing similar marks viewing them in a third
party's possession after purchase. The classic case is Levi Strauss
& Co. v. Blue Bell, Inc., 632 F.2d 817, 822 (9th Cir. 1980), in which
Levi's won an injunction against Wrangler's use of Levi's registered pocket
tabs, even though Wrangler's competing jeans were prominently labeled
at the point of sale. As articulated better in later cases, the
theory was that prospective buyers of jeans generally would see people
wearing the Wrangler jeans without the labels and take them for Levi's
jeans, thereby putting Levi's goodwill and reputation beyond its control.
See also, Polo Fashions, Inc. v. Craftex, Inc., 816 F.2d 145, 148-149
(4th Cir. 4th Cir. 1987) (imitation of plaintiff's polo player symbol
on front of shirt would cause post-sale confusion notwithstanding label
on inside of back of neck).
5. The final variety of confusion is the one at issue in Playboy
v. Netscape and Brookfield: so-called "initial interest confusion."
Unlike post-sale confusion, initial interest confusion affects that
actual purchaser of a product or service, and not a third party. The
theory is that, although the purchaser learns by the time of sale that
the trademark owner is not the source of the goods or services,
their seller has gained an opportunity to make the sale by trading on
the goodwill associated with the trademark. After using the trademark
(without authorization) to eliciti the purchaser's "initial interest,"
the seller says, in effect, "we're not the company you wanted, but we
can offer you something just as good, or maybe cheaper."
The Brookfield court cited this kind of confusion in enjoining
the use of plaintiff's trademark "MovieBuff" (with no space) as part of
defendant's domain name "moviebuff.com," and in defendant's metatags (hidden
HTML code used to attract search engines) for similar movie-related and
movie-industry-related information services. It did so although defendant's
Website was prominently marked with defendant's own name, trademarks,
and contact information. See Brookfield Communications, Inc. v.
West Coast Entertainment Corp., 174 F.3d 1036, 1057-1058, 1061, 50 U.S.P.Q.2d
(BNA) 1545 (9th Cir. 1999) (domain name); 174 F.3d at 1062-1066 (metatags).
Playboy v. Netscape, of course, involved neither domain names
nor metatags, but so-called search-engine "keys." Such a key is
a list of words which a Website can "purchase" from a search engine, so
that links to the Website appear in the list of search results whenever
the user types in one or more words on the "key" list as part
of a search request.
6. Which type of activity produces the most egregious type of initial
interest confusion and why: (1) using plaintiff's trademark as a domain
name; (2) using it as one of several or many search terms in metatags
hidden in one's own Web pages; or (3) using plaintiff's trademark as one
of the words in the list in a search-engine key? Can you articulate
reasons for your answer? As you consider these questions, consider
the major policies of trademark protection: (1) providing consumers with
an efficient "shorthand" language to identify goods and services quickly
and reliably in the marketplace; (2) avoiding consumer confusion and the
disappointment, frustration and inefficiency that it produces; (3) protecting
producers' investment in the quality and reputation of their goods and
services, thereby providing an incentive to make that investment; and
(4) preventing "dirty tricks," i.e., unfair competition, by which a rival
reaps where he has not sown. See generally, Jay Dratler, Jr., Intellectual
Property Law: Commercial, Creative and Industrial Property § 9.02
(Law Journal Press 1991, updated semiannually) (discussing trademark-protection
policies and citing authorities).
7 It is black-letter trademark law that truthful comparative
advertising is not trademark infringement, even if it relies on unauthorized
use of a rival's trademarks to make the comparison. See generally,
Dratler, supra, § 10.04. For an extreme case, see
Societe Comptoir de L`Industrie Cotonniere Etablissements Boussac v. Alexander's
Department Stores, 299 F.2d 33, 35-36 (2d Cir. 1962). Is using a
rival's trademark in a metatag or search-engine key just another form
of comparative advertising? Is it under the circumstances described
in Playboy v. Netscape? Is using a rival's trademark as your
own domain name just another form of comparative advertising?
8. Comparative advertising usually relies on the defense of trademark
"fair use," as outlined in Lanham Act § 33(b)(4), 15 U.S.C.
§ 1115(b)(4). Most courts say the defense has two elements:
(1) descriptive (non-trademark) use and (2) good faith. Which of
these elements was most at issue in Playboy v. Netscape? in
Brookfield? Did the defendant's use of "moviebuff.com"
as its own domain name support or undermine its defense of fair use with
respect to the metatags? Why?
As explained in Playboy v. Netscape, the Ninth Circuit applies
a special doctrine, called "nominative use," rather than the general doctrine
of trademark fair use, whenever confusion seems likely. The requirements
of nominative use are outlined in the court's decision at footnote
21. Is it narrower or broader than the general doctrine of fair
use? Is it consistent with the statute? with trademark policy?
9. With regard to comparison shopping, do you remember the old yellow-pages
slogan, "let your fingers to the walking?" If the telephone greatly
assisted comparison shopping (especially in inclement weather), isn't
the Web the greatest comparison-shopping tool yet invented? Some
Web browers (such as Netscape and Safari) now provide a feature known
as "tabbed browsing." See Walter
S. Mossberg, "For Tabbed Browsing and Other New Tricks,
Try Explorer's Rivals," Wall Street Journal, Thus. Jan. 8,
2004 at B1. [YOU HAVE TO HAVE A SUBSCRIPTION OR REGISTER FOR A FREE
TRIAL SUBSCRIPTION TO USE THIS LINK.] This technology allows you
(among other things) to open a series of screens containing prices, features,
and pictures of competing products simultaneously and switch among them
with a single click of your mouse. Search engines, which rely
in part on key lists and metatags, are what make this type of shopping
possible, for you have to locate the desired Websites from among tens
of millions before you can use product-specific screens like this to comparison
10. Are trademarks a valuable tool for comparison shopping? Is
using them for that purpose consistent with the policies underlying trademark
protection? In particular, is comparison shopping consistent with
the economic policy of providing a reliable, efficient "shorthand" language
for identifying products and services?
Which approach toward the use of trademarks in metatags and search-engine
keys best accomodates use of the Web for comparison shopping and, at the
same time, promotes trademark-protection policies: (1) the approach of
the Brookfield court; (2) the approach of the majority in Playboy
v. Netscape; or (3) the approach of Judge Berzon, concurring in Playboy
v. Netscape? Can you justify your answer?
11. Suppose that you were Netscape's general counsel, and that the
Ninth Circuit had just handed down its decision in Playboy v. Netscape.
What, if anything, would you advise Netscape to change in: (1) its
use of search-engine keys; (2) its restrictions on how customers can purchase
and use those keys; (3) the words used in the keys; and (4) its hit-monitoring
and pricing practices? How would you explain and justify your recommendations
to Netscape's CEO?
12. In its analysis of likelihood of confusion, the Brookfield
court made much of the fact that the defendant's domain name and metatags,
like plaintiff's trademark, had no space between the words "movie" and
"buff." See Brookfield, supra, 174 F.3d at 1049, 1066.
Indeed, that court opined that defendant could use the words "movie buff"
in its metatags, but only with the space. See id. at 1066.
Is this result appropriate, or is it legal hairsplitting? Is
it consistent with custom and usage of words on the Web? in domain
names? in metatags? Is it consistent with the law's general
approval for comparative advertising? Is it necessary to protect
trademark owners' legitimate interests?
13. For Web pages drafted in native HTML
("hypertext mark-up language"), you can view hidden metatags by using
the "View: Source" or "View: Page Source" feature of your Browser's menu.
The metatags word lists will appear after an HTML "tag" or marker
containing the word "meta". You can search for such a tag visually
or use your Browser's "Edit: Find and Replace" feature. Many Web
pages today, however, are generated on demand by scripting programs such
as PHP. You
may not be able to display the "page source" or "source HTML code" for
these pages, and so you may not be able to view their metatags.
14. In the early days of search engines, some would "rank"
search results by the number of times a particular word or phrase appeared
in the Web pages, whether in text or metatags. (Most search engines
are now more sophisticated and use other ranking techniques.) Suppose
one of Netscape's customers had put the mark "Playboy" one hundred
times in the metatags for his home page, so that, in a search for "Playboy,"
his Website appeared first, even before PEI's own and affiliated sites.
Should he be liable for trademark infringement or unfair competition
as a result of these acts? What if his metatags include the words
"nude women" one hundred times,with the same operational result?