concurring in part and dissenting in part.
I agree that the district court erred in granting summary judgment in favor of Gibson and I also agree that Gibson cannot maintain its trademark infringement claim either on a theory of likelihood of confusion at the point-of-sale (for it has disclaimed that a consumer could be confused at the point-of-sale)1 or on a theory of *554post-sale confusion.2 However, because I believe that a product shape trademark holder should be able to present evidence to maintain a trademark infringement claim on the theory of initial-interest confusion, I dissent with respect to this issue.
First, this court has recognized initial-interest confusion as an infringement under the Lanham Act. See PACCAR Inc., v. Telescan Technologies, L.L.C., 319 F.3d 243, 253 (6th Cir.2003) (citing Eli Lilly & Co. v. Natural Answers, Inc., 233 F.3d 456, 464 (7th Cir.2000) (“Such confusion, which is actionable as an infringement under the Lanham Act, occurs when a consumer is lured to a product by its similarity to a known mark, even though the consumer realizes the true identity and origin of the product before consummating a purchase.”)).3 Second, a product shape can serve as a trademark, i.e. it can identify the source of the product. See Qualitex Co. v. Jacobson Products Co., Inc., 514 U.S. 159, 172-73, 115 S.Ct. 1300, 131 L.Ed.2d 248 (1995) (recognizing that the Senate Report accompanying the Lanham act revision of 1988 stated that the “revised definition intentionally retains ... the words ‘symbol’ or ‘device’ so as not to preclude the registration of colors, shapes, sounds or configurations where they function as trademarks.”) (citation omitted and ellipsis in original); see also Kohler Co., v. Moen Inc., 12 F.3d 632, 636 (7th Cir.1993) (concluding, in a case involving faucet manufacturer, which received a trademark after demonstrating that “purchasers of its products recognized the source of its faucets by their distinctive shapes,” that product configurations were entitled to trademark protection). Since a product shape can serve as a trademark, and since initial interest confusion is a recognized theory of relief generally for trademark infringements, I believe that a product shape trademark holder whose product shape does in fact identify the product’s source should not be precluded from presenting evidence that a competitor’s product shape causes consumers to be attracted to it because of its similarity to a trademark holder’s mark.
The majority’s reason for rejecting the application of initial interest confusion to product shapes is based upon the concern that since many product shapes within the same category will appear similar when viewed from a sufficient distance, if the initial-interest confusion doctrine were applied to product shapes, a product shape trademark holder could prevent competitors from producing even dissimilar products that appeared from a sufficient distance to be somewhat similar to a *555trademarked shape. This concern, however, is misplaced. Evidence that a competitor’s product shape is similar to a trademark holder’s product shape when viewed from afar is irrelevant unless the product shape trademark holder maintains that its product shape identifies its source when viewed from afar. For if a product shape trademark holder does not assert that its product shape identifies its source when viewed from a certain distance, then any alleged confusion between the trademark holder’s product shape and a competitor’s product shape would not support the trademark holder’s claim for infringement. If a product shape trademark holder does assert that its product shape serves to identify the product’s source when viewed from a distance where many competitor products appear substantially the same, then this will be evidence that the trademark holder’s product shape does not identify its source. If most product shapes in the same product category have similar shapes, a product shape trademark holder will have a difficult time establishing that its trademark identifies the source of its product when viewed from afar, for the further one is away from a product, the more similar products in the same category will look to each other and, thus, the less likely a product shape will identify the source of the product (i.e. serve as a trademark) from that vantage point.4 In other words, a product shape trademark holder will not be able to present probative evidence of initial interest confusion unless it first shows that its product shape identifies its source when viewed from the vantage point where the confusion is alleged to have occurred.
Thus, this preliminary question (whether a product shape identifies its source when viewed from the point where the confusion is alleged to have occurred), when in dispute, as it is here, must be addressed before determining if there is an issue of fact as to whether a competitor’s product shape causes initial interest confusion due to its similarity to a trademark holder’s product shape.5 For if the product shape *556does not identify its source, then, logically, there can be no confusion as to the source of the product due to the product’s shape, since, in such a case, a consumer would not associate the product’s shape with its source.
Here, PRS argues that, as there are only a few general shapes in which guitars are made, and since there are numerous single-cutaway style guitars in the market, a consumer would not, and, indeed, could not, identify the source of a guitar on the basis of its shape. Although PRS would likely prevail on this issue, I cannot say, in light of the record as developed, that it has met its burden establishing that there is no genuine issue of material fact as to whether the shape of Gibson’s guitar identifies its source.
Since the district court incorrectly analyzed the issue of whether Gibson’s trademark served to identify its source from the distance where the confusion was alleged to have occurred, I would remand for consideration of this issue.6 If it were found that the shape of Gibson’s guitar does serve to identify its source from the distance where the confusion is alleged to have occurred, then I believe that Gibson should be entitled, just as any other trademark holder would be, to present evidence that a competitor’s use of a similar mark causes confusion as to the source of the competitor’s product.7 Thus, I respectfully dissent.
. The majority states, "Gibson’s concession that there is no point-of-sale confusion (which goes to the fourth Frisch factor, actual confusion) is dispositive of Gibson’s claim.” I believe that what is dispositive of any trademark infringement claim brought on a theory of likelihood of confusion at the point-of-sale is not Gibson’s concession of the fourth Frisch factor, which goes to evidence of actual confusion, but rather Gibson's concession that there is no likelihood of consumer confusion at the point-of-sale, which is the ultimate issue to which the Frisch factors are used as a guide to help address when analyzing a trademark infringement claim brought under a theory of likelihood of confusion at the point-of-sale.
. The last sentence of the post-sale confusion section reads, "[P]ost-sale confusion cannot serve as a substitute for point-of-sale confusion in this case." I view the post-sale confusion theory of relief not, when applicable, as serving as a substitute for point-of-sale confusion, but rather as a distinct theory of relief. When the alleged consumer confusion takes place at the point-of-sale, the ultimate issue is likelihood of confusion at the point-of-sale, in which the plaintiff will introduce evidence of point-of-sale confusion. However, when the alleged consumer confusion takes place in the marketplace (post-sale), then the ultimate issue is likelihood of post-sale confusion, in which the plaintiff introduces evidence of post-sale confusion.
. The majority, in footnote 15, explains that PACCAR’s discussion of initial interest confusion was specifically in reference to a dispute over internet-domain names. Obviously, as applied in that case, the doctrine's application related to a company’s name trademark as used in an internet-domain name. All trademarks, whether they be a company’s name, logo, or a product shape, do the same thing: they identify the source of a good. Since we (and many other circuits) have recognized the doctrine, the majority must explain why this recognized doctrine should not apply to trademarked product shapes. I find the majority's explanation unpersuasive.
. The majority repeats its concern underlying its rationale for rejecting the application of initial interest confusion to product shapes in footnote 15 by noting that since many competing products will create initial interest due to their resemblance with a trademarked product shape when viewed from afar, a trademark product shape holder would be able to protect a "penumbra” of more or less similar shapes. If there exists a "penumbra” of products in the marketplace with shapes similar to a trademarked product shape when viewed from afar, then the trademarked product shape would likely fail to identify its source. If there does not exist a "penumbra” of products with shapes similar to a trademarked product shape when viewed from afar, and if the product shape does identify its source when viewed from that vantage point, and if the functionality doctrine does not prevent the application of the initial interest confusion doctrine when such confusion is alleged, then I believe a product shape trademark holder should be able to maintain a trademark infringement claim on the theory of initial interest confusion.
. The majority views this inquiry as "needlessly complicated and unworkable.” I view it as essential. Since, as noted before, the purpose of every trademark is to identify the source of its good, if a trademark does not identify its source, a trademark holder cannot successfully prosecute a trademark infringement claim. If a product shape trademark holder believes that a competitor's product causes confusion as to its source due to the competitor’s products resemblance to the trademark holder’s product shape when viewed from a store's aisle or front-door, then the product shape trademark holder will need to establish that its trademark identifies its source when viewed from those perspectives for any evidence of such confusion to support its claim. Thus, the relevant observation distance or location for determining whether the trademark holder's product shape identifies its source is from whatever distance or loca*556tion the trademark holder alleges confusion occurred.
It is true, of course, that a product shape trademark holder could argue that a competitor's product causes confusion as to its source when viewed from varying perspectives. Assuming the defendant argues that the trademark holder's product shape does not identify its source when viewed from any of those perspectives, the court will need to address this argument. In this case, the only relevant allegation of initial-interest confusion is from Gibson’s expert, who claimed that upon entering Gruhn's Guitar store in Nashville, he looked at what he thought was a wall of Gibson Les Paul guitars to discover, upon closer inspection, that the guitars were actually PRS Singlecuts. Since PRS argues that Gibson’s guitar shape does not identify its source, Gibson must establish that its shape does identify its source from the perspective where the alleged confusion occurred. Ultimately, Gibson may not be able to succeed in establishing this point, but PRS has not met its burden to be entitled to summary judgment.
. The district court made two fundamental errors with respect to its analysis of this issue. First, the district court improperly considered factors beyond the shape of Gibson's guitar, such as knob placement and color, in determining that Gibson's trademark identified its source, since Gibson's trademark consists only of the shape of the guitar. And second, the district court improperly concluded that since evidence of widespread use of similar marks "is not sufficient to rebut as a matter of law the presumption” that an incontestable mark identifies its source, such evidence "does not warrant further consideration.” 311 F.Supp.2d 690, 718. Rather, this evidence, which goes to the heart of the issue of whether Gibson's shape can identify its source, deserved significant consideration, even if such evidence were found not to be sufficient as a matter of law to establish that Gibson's mark does not identify its source.
. If the initial interest confusion doctrine were not applicable to product shape trademarks, I believe that product shape trademark holders would be put to a significant disadvantage compared to non-product shape trademark holders since this type of confusion is likely the only type of confusion that could arise with product shapes. It will be the rare case for there to exist a likelihood of confusion with respect to product shape trademarks at the point of sale since any confusion created by a product's similar shape will be dispelled at the point of sale, as most products will have a distinguishing feature that is *557identifiable from this vantage point, such as its producer’s name. Thus, if initial interest confusion were not a viable theory upon which a product shape trademark holder could proceed, a product shape trademark holder may be quite limited, if not foreclosed, from successfully prosecuting a trademark infringement claim.