SHADUR, District Judge:
The Morningside Group Limited (“Morningside Group”) brought suit against Morningside Capital Group, L.L.C., (“Morningside Capital”) for unfair competition and trademark dilution under Lanham Act §§ 43(a) and (c)(15 U.S.C. 1125(a) and (c)), in addition to various state law claims. After a three-day bench trial in February 1998, the district court (Janet C. Hall, Judge) entered judgment for Morningside Capital, having found that Morningside Group’s name did not constitute a service mark and that even if it did, it was not infringed because confusion was not likely.
Morningside Group appeals that judgment of dismissal, raising two principal issues: (1) whether the district court correctly held that Morningside Group does not provide services to others and therefore does not have a service mark entitled to protection under the Lanham Act; and (2) if a service mark is at issue, whether the district court correctly found that Morningside Capital does not infringe that mark because no likelihood of confusion exists under the factors articulated in Judge Friendly’s opinion for this court in
Polaroid Corp. v. Polarad Electronics Corp.,
287 F.2d 492, 495 (2d Cir.1961). We reverse on both grounds.
Background
We begin with a brief overview of the history and operations of the two financial institutions at war in this case. We will then turn to a brief recapitulation of the lawsuit itself.
Morningside Group
In the late 1980s members of the Hong-Kong-based Chan family began building a financial investment business. At that time they selected the “Morningside” mark to identify their evolving corporation and its affiliates, subsidiaries and licensees. Their primary corporation is The Morning-side Group Limited, the plaintiff in this action and a British Virgin Islands corporation. According to the Chans, they chose the name “Morningside” because of the positive connotations of the word “morning” in Chinese, and because the first member of their family to be educated in the United States had attended Columbia University in New York City’s Morn-ingside Heights neighborhood.
Together the Morningside Group businesses and their licensees have engaged in various financial activities in the United States. For example, as the district court explained, members of the Morningside Group have structured and financed investments in a broad range of United States companies and assisted in managing their acquisitions. Significantly (though the district court viewed this as unimportant) Morningside Group has often solicited co-investors in those endeavors.
Members of the Morningside Group have also assisted Asian companies in acquiring United States technology through negotiations of joint ventures and licensing transactions. Finally, Morningside (China) has provided financial advice to United States investors regarding potential investments in Asia.
Morningside Capital
Morningside Capital is a Connecticut company formed in 1995 by business executive Vincent Wasik (‘Wasik”) and corporate lawyer Laurence Bathgate (“Bath-gate”). Wasik and Bathgate chose the name “Morningside” because their offices are located at One Morningside Drive in Westport, Conn. Morningside Capital is a private equity investment group that sponsors, negotiates and finances the acquisitions of companies.
Events Leading to This Laiosuit
Morningside Capital first generated meaningful publicity in October 1995 when it advertised its recent acquisition of Carson Products Company (“Carson Products”) in the
Wall Street Journal.
In that advertisement it identified itself as Morn-ingside Capital Group of Westport, Conn. That morning Jacullo (whose Morningside Group offices are also in Westport) received several phone calls from financial professionals congratulating him on the Carson Products acquisition or inquiring as to why they had not been informed of the deal. Since then Jacullo has continued to receive phone calls meant for Morning-side Capital and to field questions about the relationship between the two entities.
In response to the
Wall Street Journal
advertisement and the ensuing inquiries about Morningside Capital, Morningside Group filed a Lanham Act suit in the District of Connecticut. Ultimately a bench trial was held, and in July 1998 the district court entered judgment in favor of Morningside Capital. This is the appeal from that judgment.
Standard of Review
We review for clear error whether an entity provides a service and whether a mark has been used to identify a particular service, because both of those determinations are generally findings of fact (see, e.g.,
In re Advertising & Marketing Dev., Inc.,
821 F.2d 614, 621 (Fed.Cir.1987)). As both we and the Supreme Court have noted in other contexts, however, factual findings based on faulty legal analysis are not insulated by the clearly erroneous standard (see
Inwood Labs., Inc. v. Ives Labs., Inc.,
456 U.S. 844, 855 n. 15, 102 S.Ct. 2182, 72 L.Ed.2d 606 (1982);
Thomson v. Larson,
147 F.3d 195, 199 (2d Cir.1998)). That principle applies here, and we keep it in mind as we review the district court’s determination that Morningside Group does not have a valid service mark.
We will then turn to the district court’s conclusion that there was not a likelihood of confusion under the
Polaroid
factors. In that respect, while a district-court’s finding of fact as to each
Polaroid
factor is reversed only if clearly erroneous, “[t]he ultimate weighing of the factors ... is reviewed de novo”
(Estee Lauder Inc. v. The Gap, Inc.,
108 F.3d 1503, 1510 (2d Cir.1997)).
“Morningside Group” as a Service Mark
To prevail on a Lanham Act infringement claim, a claimant must show that “it has a valid mark entitled to protection and that the defendant’s use of it is likely to cause confusion”
(Cadbury Beverages, Inc. v. Cott Corp.,
73 F.3d 474, 477 (2d Cir.1996), quoting
Gruner + Jahr USA Publishing v. Meredith Corp.,
991 F.2d 1072, 1075 (2d Cir.1993)). We therefore begin our analysis by reviewing the district court’s initial determination": that “Morningside Group” is not a valid service mark because plaintiff does not provide “services” within the meaning of the Lanham Act.
Lanham Act § 43(a) states in relevant part (emphasis added):
(1) Any person who, on or in connection with any goods or
services, ...
uses in commerce any word, term, name, symbol, or device, or any combination thereof, ... which—
(A) is likely to cause confusion, or to cause mistake, or to deceive as to the affiliation, connection, or association of such person with another person, or as to the origin, sponsorship, or approval of his or her goods,
services,
or commercial activities by another person,
ífc >¡i jJ; í¡í #
shall be liable in a civil action by any person who believes that he or she is or is likely to be damaged by such act.
“Services” as used in the Lanham Act has been defined as “the performance of labor for the benefit of another”
(Advertising & Marketing,
821 F.2d at 619, quoting
In re
Canadian Pac. Ltd.,
754 F.2d 992, 994 (Fed.Cir.1985)). Those services must not be “solely for the benefit of the performer; the services must be rendered to others”
(Murphy v. Provident Mut. Life Ins. Co.,
923 F.2d 923, 927 (2d Cir.1990)). Further, the claimant must have used the claimed mark to identify its services
(Advertising & Marketing,
821 F.2d at 620).
Morningside Group’s activities clearly fall within that definition. In concluding otherwise the district court relied on an incorrect legal standard:
It found no services because Morningside Group’s activities “were all intended to benefit plaintiff and the Chans alone,” not third parties. That analysis is mistaken because the definition of “services” does not require that the service provider be
motivated
by a desire to benefit others — ’indeed, the goal of most service providers (unsurprisingly in a free enterprise economy) is ultimately to benefit themselves. Instead the issue is whether the service provider
in fact
benefits third parties, regardless of its reason for providing its services.
Morningside Group has done just that. Its primary motivation — indeed, perhaps its only ultimate motivation — may well have been to benefit itself and the Chan family. But its activities did provide valuable benefits to others, as the district court recognized. As long as those “others” are truly third parties and not simply branches or affiliates of Morningside Group, Morningside Group is providing “services” within the meaning of the Lanham Act (see
Canadian Pacific,
754 F.2d at 995-96). Morningside Group has demonstrated that it renders services to others — including, for example, the businesses they acquire, the co-investors they attract and the United States institutions whose money they direct to Asian investments.
Furthermore, this is not a case like
Buti v. Impressa Perosa, S.R.L.,
139 F.3d 98 (2d Cir.1998), in which the service mark claimant did not provide its actual business services (the conduct of a restaurant business) in the United States, but only advertised and promoted those services in the United States. Mere advertising and promotion of a mark in this country are not enough to constitute “use” of the mark “in commerce,” so as to bring the activity within the scope of the Lanham Act
(id.
at 105). By contrast, here Morningside Group does conduct material aspects of its investment services in the United States.
Because Morningside Group provides “services” within the meaning of the Lan-ham Act, its mark is a protectible service mark. To determine whether Morningside Capital has infringed that mark, we turn to the likelihood-of-confusion analysis.
Likelihood of Confusion: Polaroid Factors
Once a claimant has demonstrated that it has a valid mark, it must then prove that “numerous ordinary prudent purchasers are likely to be misled or confused as to the source of the product in question because of the entrance in the marketplace of defendant’s mark”
(Cadbury Beverages,
73 F.3d at 477-78, again quoting
Gruner
+
Jahr,
991 F.2d at 1077).
Polaroid,
287 F.2d at 495 identifies eight factors relevant to that likelihood of confusion inquiry:
1. strength of plaintiffs mark;
2. degree of similarity between the two marks;
3. proximity of the products (or services);
4. likelihood that the prior owner will “bridge the gap”;
5. evidence of actual confusion;
6. defendant’s good faith in adopting its mark;
7. quality of defendant’s product (or services); and
8. sophistication of the buyers,
No one of those factors is dispositive, and the list is not exhaustive (id). We discuss the listed factors in turn.
Strength of Plaintiffs Mark
As to the first factor, the district court found Morningside Group’s mark to be weak. Although the district court initially (and correctly) characterized the mark as arbitrary, it went on to find the mark weak because of third-party use of the “Morn-ingside” mark for unrelated types of services and because of Morningside Group’s lack of advertising.
When a mark has been deemed arbitrary, as opposed to generic, descriptive or suggestive, its very arbitrariness is an indicium of strength
(Sports Auth., Inc. v. Prime Hospitality Corp.,
89 F.3d 955, 961 (2d Cir.1996)). Importantly, a mark’s strength is examined principally in the market in which the mark is used. Hence defining that relevant market becomes an important aspect of the infringement analysis. Here the relevant market is the relatively small world of financial investment professionals. But the district court did not limit itself to that market in its examination of third-party users of the “Morningside” mark- — a faulty analysis that led to the ultimately incorrect conclusion that the mark is weak.
Use of a like mark in a different market for different products or services need not undermine the mark’s strength in its own market
(Scarves by Vera, Inc. v. Todo Imports Ltd.,
544 F.2d 1167, 1173-74 (2d Cir.1976); see also
Century 21 Real Estate Corp. v. Century Life of Am.,
970 F.2d 874, 877-78 (Fed.Cir.1992)). In this instance the district court looked to businesses that used “Morning-side” in their titles but were not shown to be using the mark within the relevant market. Oddly enough, it cited as an example of such third-party use Morningside Capital Management, an Atlanta institutional money management firm that agreed to
discontinue
its use of the Morningside mark after being sued by Morningside Group for trademark infringement. That example does not weaken the mark — instead it strengthens it, because the successful policing of a mark adds to its strength to the extent that it prevents weakening of the mark’s distinctiveness in the relevant market (see, e.g.,
Lexington Management Corp. v. Lexington Capital Partners,
10 F.Supp.2d 271, 283 (S.D.N.Y.1998)).
Finally, the district court improperly relied on Morningside Group’s lack of advertising as evidence of the mark’s weakness. Although evidence of advertising can bolster a mark’s strength, its absence need not weaken the mark
(McGregor-Doniger Inc. v. Drizzle Inc.,
599 F.2d 1126, 1132 (2d Cir.1979)),
particularly when, as the district court found to be the case here, “the people wishing to reach plaintiffs licensees already know who they are before they attempt the contact.” When a claimant has no need for traditional advertising because of the nature of its market, it should not feel compelled to advertise simply to protect its service mark.
While we do not dispute the factual underpinnings for the district court’s conclusion that the mark is weak, we find that the correct analysis of those facts clearly calls for the opposite conclusion. Because there was no evidence that third-party use of similar marks in the relevant market undermined the inherent distinctiveness of plaintiffs mark and because the mark seems to have been somewhat well known in that market, we hold the mark is relatively strong.
Degree of Similarity Between the Marks
In assessing the similarity between two marks, a court will look to whether the similarity is likely to cause
consumer confusion
(Sports Auth.,
89 F.3d at 962). But as we have said in
Estee Lauder,
108 F.3d at 1511:
The test, however, is not whether confusion is possible; nor is it whether confusion is probable among customers who are not knowledgeable. Rather, the test ... is whether confusion is probable among numerous customers who are ordinarily prudent.
We disagree with the district court’s determination that the two marks do not “convey the same overall impression,” leading it to the mistaken conclusion that the factor favors Morningside Capital.
We find unpersuasive any notion that the two marks are dissimilar. Consider again the full names of the two entities: “The Morningside Group Limited” and “Morningside Capital Group, L.L.C.” Only one word stands out as dominant in the two names, and that—the only arbitrary word—is “Morningside.” Both entities use the word “Group” in their names, and the only other word in defendant’s name (apart from the general corporate designation “L.L.C.,” which finds its echo in the word “Limited” in plaintiffs name) does not serve any differentiating role.
Considering “the general impression created by the marks, taking into account all factors that potential purchasers will likely perceive and remember”
(Lang v. Retirement Living Publishing Co.,
949 F.2d 576, 581 (2d Cir.1991)), we find it clear that the similarity between the two names would be likely to cause confusion among ordinarily prudent consumers in the investment world. This factor too favors plaintiff.
Proximity of Services
This third
Polaroid
factor “concerns whether and to what extent the two products compete with each other” and “the nature of the products themselves and the structure of the relevant market”
(Cadbury Beverages,
73 F.3d at 480). Certainly the parties here are in competition. As recognized by the district court, “they both seek out direct investment opportunities for themselves and others.” Although the two entities had focused then-investment efforts on somewhat different industries, their services need not be identical
(Arrow Fastener Co. v. Stanley Works,
59 F.3d 384, 396 (2d Cir.1995);
Lexington Management,
10 F.Supp.2d at 284-85). When they provide essentially the same service to the same customer base, their services “are related and proximate” (in the words of the district court).
Furthermore, the nature of the financial investment market renders the proximity of the two company’s services particularly troubling in likelihood-of-confusion terms. High-level investment business is commonly conducted based on trust and personal relationships among individuals. If any investor reads about a Morningside Capital investment—such as the Carson Products acquisition'—and mistakes that transaction for a Morningside Group investment of which it was not aware, it may well feel “cut out” of a potentially lucrative deal.
Unless it communicates with someone at Morningside Group to express its displeasure and thus discovers its mistake, its business relations with Morningside Group could be soured. In that market, then, even positive publicity for Morningside Capital could cause business problems and lost opportunities for Morningside Group. And it is of course obvious that adverse consequences for Morningside Group could also flow from a converse scenario—unfavorable publicity attendant on a Morning-side Capital transaction.
Indeed, as stated in a different context, there may be “no effective way to measure the loss of sales or potential growth — to ascertain the people who don’t knock on the door or to identify the specific persons who do not [return] because of the existence of the infringer”
(Hyatt Corp. v. Hyatt Legal Services,
736 F.2d 1153, 1158 (7th Cir.1984) (citation omitted)). Direct competition between the parties, when coupled with the personal nature of the investment market, makes it likely both that confusion would occur and that it would harm Morningside Group. We conclude, as did the district court, that the proximity factor weighs heavily in favor of Morningside Group.
Likelihood That the Prior Owner Will “Bridge the Gap”
This factor inquires whether a plaintiff is likely to enter defendant’s market, or “bridge the gap.” It recognizes “the senior user’s interest in preserving avenues of expansion and entering into related fields”
(Hormel Foods Corp. v. Jim Henson Productions, Inc.,
73 F.3d 497, 504 (2d Cir.1996) (citation omitted)). We agree with the district court that given the close proximity of the services provided by the two entities, plus Morningside Group’s interest in the fields in which Morningside Capital invests, this factor also favors Morningside Group.
Evidence of Actual Confusion
Evidence that confusion has actually occurred is of course convincing evidence that confusion is likely to occur (see, e.g.,
Streetwise Maps, Inc. v. Vandam, Inc.,
159 F.3d 739, 745 (2d Cir.1998)). At trial Morningside Group presented extensive evidence of phone calls and other inquiries received by its people from sophisticated members of the financial community, both about Morningside Capital transactions and about the relationship between the two entities.
Nonetheless the district court discounted that evidence because it relied on an inordinately narrow definition of actual confusion. Contrary to the district court’s approach, evidence of actual confusion need not be limited to evidence of mistaken
completed
transactions. Nor is the inquiry confined to evidence that Morning-side Capital was able to “pass off’ its services as those of Morningside Group. Instead, evidence of actual confusion regarding affiliation or sponsorship is also entirely relevant to the ultimate likelihood-of-confusion inquiry.
In fact, Lanham Act § 43(a)(1)(A) itself states that infringement occurs when defendant’s actions are “likely to deceive as to the affiliation, connection, or association” of the defendant with the plaintiff. As we have reconfirmed in
Sports Auth.,
89 F.3d at 963, courts can properly take into account evidence of “either a diversion of sales, damage to goodwill, or loss of control over reputation.” There
(id.
at 963-64) we found important to the actual confusion factor the evidence that consumers placed phone calls to the wrong party and mistakenly believed that plaintiffs and defendant’s services were affiliated because of their similar names. That is precisely the type of evidence presented by Morningside Group.
Recognizing the relevance of actual confusion beyond mistaken completed transactions is important here, because in the financial world an investor will almost never complete a transaction with a mistakenly identified party. If nothing else, compliance with the due diligence requirement will normally prevent such errors. But investors might be confused about the affiliation between two similarly named companies and might very well alter their behavior based on that confusion. Here Morningside Group has demonstrated just such confusion by its customers, particularly with regard to Morningside Group’s role in Morningside Capital’s acquisition of Carson Products.
In sum, Morningside Group has demonstrated actual confusion as to sponsorship or affiliation, and such actual confusion is
often the most telling indication that confusion is likely. Hence the' actual confusion factor also weighs heavily in Morningside Group’s, favor. .
Defendant’s Good Faith
We agree with the district court’s conclusion that Morningside Capital acted in good faith when it adopted its name. Morningside Capital’s principals chose the name because of their office address at One Morningside Drive, and they were not aware of Morningside Group when they did so. Further, they conducted a corporate name search in Connecticut and Delaware and did not come across Morningside Group. It is irrelevant that they did not also conduct a trademark search, for Morningside Group had not yet registered its Morningside mark when Morningside Capital’s name was chosen. Thus this factor favors Morningside Capital, but it is obvious that because the ultimate issue is likelihood of confusion, a defendant’s good faith alone will not preclude a finding of infringement where such a finding is dictated by the remaining
Polaroid
factors.
Quality of Defendant’s Services
Under this factor a court first examines whether defendant’s products or services are inferior to plaintiffs, thereby tarnishing plaintiffs reputation if consumers confuse the two. Noting the excellent reputations of both Morningside Group and Morningside Capital, the district court determined that this factor favors Morn-ingside Capital.
We cannot say that the district court was clearly erroneous in that assessment. Except for one brief criticism of Carson Products (a Morningside Capital acquisition) in a generally positive
Fortune
magazine article (see n. 7), Morningside Group has not demonstrated that _ Morningside Capital provides inferior financial services.
But having said that, we note that the district court did not examine the second way in which the quality of a defendant’s services can be relevant: Products of equal quality may tend to create confusion as to source because of that very similarity of quality
(Hormel Foods,
73 F.3d at 505). As the district court said, the evidence seems to indicate that the companies provide comparable services. But it is precisely that similarity that might prompt a Morningside Group customer to mistake a publicized Morningside Capital transaction for a Morningside Group deal. And as we have said, that confusion can injure Morningside Group’s reputation not because a consumer mistook an inferior Morningside Capital transaction for a Morningside Group transaction, but because a Morningside Group co-investor is upsfet about not being invited to participate in the deal—something that has actually happened (see n. 6).
In a sense, then, this factor cuts both ways. True enough, Morningside Capital does not provide inferior services and so does not harm Morningside Group’s reputation in that way. But the comparable quality of the two companies’ services brings them into even closer proximity, thereby militating in favor of Morningside Group.
Sophistication of the Buyers
We turn finally to the last
Polaroid
factor: the sophistication of the relevant purchasers. There is no dispute that the customers for the services of both Morningside Group and Morningside Capital are sophisticated. As the district court found, they are a small circle of professionals, comprising “investment bankers, private investors, accountants, attorneys and business owners and managers.” They are often able to invest large sums of money and do so only after extensive research. Clearly they constitute a highly sophisticated market.
But the district court erred when it woodenly applied the general rule that likelihood of confusion is diminished when potential purchasers are sophisticated (see, e.g.,
Paddington Corp. v. Attiki Importers & Distribs., Inc.,
996 F.2d 577, 587 (2d
Cir.1993)). Indeed,
Centaur Communications, Ltd. v. A/S/M Communications, Inc.,
830 F.2d 1217, 1228 (2d Cir.1987) has noted that market sophistication might in some circumstances increase the likelihood of confusion.
But we need not rely on speculation as to the effect of the con-cededly small and highly sophisticated market here. It is clear in all events that when, as here, there is a high degree of similarity between the parties’ services and marks, “the sophistication of the buyers cannot be relied on to prevent confusion” (McGregor-Doniger; 599 F.2d at 1137). And in this instance Morningside Group presented evidence of actual confusion among several highly sophisticated investment professionals. Accordingly, in the circumstances presented, the fact of such sophistication among the relevant consumers is of no help to Morningside Capital.
Balancing the Polaroid Factors
Taking these factors together, we determine as a matter of law that they unquestionably create a likelihood of confusion among prospective purchasers of Morning-side Group’s services. Of the eight
Polaroid
factors, only one — and one of little weight in the circumstances presented— favors Morningside Capital, while a second may be viewed as favoring either litigant. All of the other
Polaroid
factors are either plainly on Morningside Group’s side of the balance scales or, as to the sophistication of the consumer group, of no help to Morn-ingside Capital.
Of all the factors, Morningside Group’s evidence of actual confusion is perhaps most probative of the likelihood that confusion will continue to occur. Indeed, where as here some members of the sophisticated market in which the parties operate have already been
actually
confused, it cannot be gainsaid that others in the same market are
likely
to be confused. Furthermore, the similarity of the marks and the proximity (both in kind and in quality) of the services provided by Morningside Group and Morningside Capital create a greater likelihood of confusion, as does the arbitrariness of the Morningside mark.
Lined up against that overwhelming evidence of likely confusion is the lone factor that favors Morningside Capital: Its name was chosen in good faith. But as already discussed, that is simply not enough for defendant to prevail.
We therefore find that Morningside Group has a protectible service mark and that its mark has been infringed because of the likelihood that consumers will be confused by Morningside Capital’s name. We reverse the district court on that claim and remand with directions to the district court to enter a permanent injunction barring Morningside Capital from using any Morningside mark and to determine what further relief might be warranted on that claim.
Dilution Claim
Morningside Group has also asserted a federal dilution claim against Morningside Capital under Lanham Act § 43(c)(15 U.S.C. § 1125(c)). That claim too was dismissed by the district court. Although it is open to substantial question whether Morningside Group could prevail on such a dilution claim (is “Morningside Group” really a “famous mark” for statutory purposes?), we need not address the subject because Morningside Group — having already succeeded on its infringement' claim — has neither requested, nor could it receive, any further relief based on dilution.
State Laiv Claims
Morningside Group also brought four state law claims, as supplemental to its Lanham Act claims: (1) injury to business
reputation and dilution under Conn. Gen. Stat. § 35 — lli(c), (2) unfair trade practices under Conn. Gen.Stat. § 42-110b(a), (3) trademark infringement under Connecticut common law and (4) unfair competition under Connecticut common law. Each of those claims was dismissed as well.
Morningside Group does not appeal the dismissal of the first state statutory claim for injury to business reputation and dilution under Conn. Gen.Stat. § 35 — lli(c). As for the two state common law claims, those also need not be addressed because Morningside Group has requested no relief under Connecticut common law beyond what it will obtain under its successful Lanham Act claim.
But the state statutory claim for unfair trade practices under Conn. Gen. Stat. § 42-110b must be remanded to the district court together with the Lanham Act claim. Morningside Group has requested different relief for that potential state law violation than for its federal infringement claim: It has asked for punitive damages under the state statute and for treble damages under the Lanham Act. We leave it to the district court to determine whether Morningside Capital has violated that state statute and whether any additional monetary relief might be permitted or warranted under that provision.
Conclusion
We ReveRse the district court’s dismissal of Morningside Group’s Lanham Act claim for infringement and RemaNd for the entry of a permanent injunction against Morningside Capital’s use of any “Morn-ingside” mark and for any other relief that may be warranted in light of this opinion. We also Reverse the dismissal of Morning-side Group’s state statutory claim under Conn. GemStat. § 42-110b and Remand that claim for a determination of Morning-side Capital’s liability and further damages, if any.