OPINION OF THE COURT

ROTH, Circuit Judge:

This appeal arises from a trademark dispute regarding Mars Sales Company’s registered trademark, “Air Door.” Berner International Corporation (Berner) filed the present action for declaratory judgment, seeking to establish that “Air Door” is a generic or common descriptive name of Mars’ product and for that reason not subject to exclusive use. The suit also seeks cancellation of Mars’ federal trademark registration of “Air Door” and an injunction prohibiting Mars from threatening Berner or its customers with claims of trademark infringement for using the term “air door” to identify or describe their products. Berner filed a motion for summary judgment, limited solely to the issue of whether Mars’ Air Door trademark was generic when it was adopted by Mars in October 1962. The district court found that “Air Door” was generic at that time; it ordered Mars’ trademark registration invalidated and permanently enjoined Mars from threatening to sue Berner for trade-fnark infringement for use of the phrase “Air Door.” Because we conclude that the district court utilized incorrect legal criteria in its analysis of whether “Air Door” was a generic term in 1962, we will reverse and *977 remand this case to the district court for further proceedings consistent with this opinion.

Mars and Berner compete in the production and sale of devices that use a barrier of moving air directed across an open doorway or other opening for temperature and insect control. The parties agree that one generic appellation for such devices is “air curtain.” Air curtains were initially manufactured in the United Kingdom and sold by the Dyfoam Corporation. In 1960 or 1961, the Miniveil division of Dyfoam was organized to market the product in the United States. In 1966, Miniveil was separately incorporated as the Miniveil Corporation; its name was changed in 1974 to Berner International Corporation. Berner initially marketed its product under the “Miniveil” trademark designation.

Mars began selling air curtains in 1962. Mars alleges that Martin Smilo, president of Mars, coined and adopted the term “air door” to distinguish his device from others on the market. Importantly, Mars alleges that in 1962, when it first adopted the mark, “air door” was not commonly used by the relevant consumers of its products; therefore, Mars contends that “air door” was not “generic” when it was adopted as a trademark. Mars alleges a continual use of the term as a trademark on and in connection with its product from the time of its entrance into the “air curtain” market until the present. Mars has advertised its product extensively under the “air door” mark and since 1980 has conducted business under the name “MARS AIR DOORS.”

When Mars first entered the market, its product was relatively new in this country and was referred to by a number of names with “air curtain” being the most common. However, the terms “air door” and “air curtain door” were also used in the manufacturing industry to some extent to describe this product. Mars alleges that, until very recently, it was the only manufacturer to use the term “air door” as a designation for its product.

During the 1970’s and 1980’s, a number of manufacturers joined and left the industry. Smilo has stated that only in the last few years has he been aware that companies other than Mars have begun to use the term “air door” to describe their products. When another manufacturer, Leading Edge, used the term in connection with or on its product, Mars objected and the company refrained from such use. Early in 1989, Mars learned that Berner was using the term “air door” in promotion of its products. After objections by Mars, Ber-ner filed this lawsuit seeking a declaratory judgment.

In support of its position that “air door” is not generic, Mars has offered the testimony of Martin Smilo, its founder and president, of Herbert Edey, president of Edey Manufacturing Company, a manufacturer and seller of air curtains from 1961 to 1976, and of Jack Haugen, an individual involved in the air curtain business since 1961. All of these witnesses assert that “air curtain” was and has been the only commonly used appellation for their products. Mars contends that the term “air door” was essentially invented by Martin Smilo in 1962 and in any event was not commonly used at that time. Mars also has introduced into evidence numerous brochures, instruction sheets, announcements and similar documents that purportedly demonstrate that members both of the air curtain industry and of the consuming public generally referred to the products at issue as “air curtains” or other designations that did not include “air door.” Finally, Mars has submitted a number of articles from trade journals and patents from the early 1960’s that use terms other than “air door” to refer to these products.

Berner, on the other hand, alleges that throughout the period from initial sales of these products in the United States to and after 1962, the products sold by Dyfoam and Miniveil were referred to interchangeably as “air curtains” or “air doors.” Ber-ner contends that the term “air door” was used as an explanation to customers to describe the device, i.e., “a door comprised of a moving screen or curtain of air, in *978 other words an ‘air door.’ ” Berner asserts that both “air door” and “air curtain” were generic or common descriptive names for this type of product during the period from 1957 to 1962.

Berner, in support of its position, has offered the testimony of Erling Berner, the Founder of Dyfoam, Miniveil and now Ber-ner Corporation, who asserts that both “air curtain” and “air door” were common descriptive names for his product during the period from 1957 to 1962. Moreover, Ber-ner has introduced four trade journal articles from the relevant time period that refer to the products at issue as “air doors” and has offered a number of its own business records from 1960-62 that also refer to these products as “air doors.” Berner has also offered Mars publicity articles that Berner contends use the term “air door” generically. Finally, Berner offers one dictionary from 1963 that defines “air door” as “[a] door for controlling air currents in a mine.”

II.

The district court had jurisdiction over this Lanham Act case pursuant to 28 U.S.C. §§ 1331 and 1338. This court has appellate jurisdiction over the final order of the district court pursuant to 28 U.S.C. § 1291.

This court’s review of a grant of summary judgment is plenary. See Erie Telecommunications, Inc. v. City of Erie, 853 F.2d 1084, 1093 (3d Cir.1988). “On review the appellate court is required to apply the same test the district court should have utilized initially.” Goodman v. Mead Johnson & Co., 534 F.2d 566, 573 (3d Cir.1976), ce rt. denied, 429 U.S. 1038, 97 S.Ct. 732, 50 L.Ed.2d 748 (1977).

A court may grant summary judgment only when the submissions in the record “show that there is no genuine issue as to any material fact and that the moving party is entitled to judgment as a matter of law.” Fed.R.Civ.P. 56(c). “The inquiry performed is the threshold inquiry of determining whether there is the need of a trial — whether, in other words, there are any genuine factual issues that properly can be resolved only by a finder of fact because they may reasonably be resolved in favor of either party.” Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 250, 106 S.Ct. 2505, 2511, 91 L.Ed.2d 202 (1986). The party opposing summary judgment must “do more than simply show that there is some metaphysical doubt as to material facts.” Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 586, 106 S.Ct. 1348, 1356, 89 L.Ed.2d 538 (1986). If the opposing party’s “evidence is merely colorable or is not significantly probative, summary judgment may be granted.” Anderson, 477 U.S..at 249-50, 106 S.Ct. at 2510-11 (citations omitted).

However, in determining whether summary judgment is appropriate, “[t]he evidence of the non-movant is to be believed, and all justifiable inferences are to be drawn in his favor.” Id. at 255, 106 S.Ct. at 2513. The inquiry is “whether the evidence presents a sufficient disagreement to require submission to a jury or whether it is so one-sided that one party must prevail as a matter of law.” Id. at 251-52, 106 S.Ct. at 2512. If the district court failed “to apply the correct legal standard” to its threshold factual inquiry, we must reverse or remand for reevaluation of the record. See A.J. Canfield Co. v. Honickman, 808 F.2d 291, 307 (3d Cir.1986).

III.

In the present case, Berner sought a declaratory judgment that the term “air door” is a generic term and was so at the time Mars adopted the mark in 1962; Ber-ner also sought an injunction prohibiting Mars from threatening to sue Berner for its use of the term. The motion for summary judgment was limited to the issue of whether the term “air door” was generic in 1962 when it was first adopted by Mars. 1 The district court found that the term was *979 generic at that time and, therefore, cannot properly be the subject of trademark protection. See Berner Int’l Corp. v. Mars Sales Co., No. 89-1999 (W.D.Pa. July 15, 1991) [hereinafter “district court opinion”], at 7-13. The court held that “members of the manufacturing industry” commonly used the term “air door” prior to 1962 and, therefore, that no genuine issue of material fact existed regarding whether the term was generic and proscribed from Lanham Act protection. See id. at 9-13. Appellant contends that the district court erred in granting summary judgment for Berner because it applied an incorrect legal standard to determine whether the term “air door” was generic in 1962 and because, even under the standard employed by the district court, there was a genuine issue of material fact. Appellant also contends that the district court improperly excluded evidence relevant to its determination. We agree that the district court applied an incorrect legal standard in its determination and will reverse and remand for proceedings consistent with this opinion.

The Lanham Act, 15 U.S.C. § 1051 et seq., provides national protection for trademarks used in interstate and foreign commerce. See S.Rep. No. 1333, 79th Cong., 2d Sess. 5 (1946). “Because trademarks desirably promote competition and the maintenance of product quality, Congress determined that ‘a sound public policy requires that trademarks should receive nationally the greatest protection that can be given them.’ ” Park ‘N Fly, Inc. v. Dollar Park and Fly, Inc., 469 U.S. 189, 193, 105 S.Ct. 658, 661, 83 L.Ed.2d 582 (1985) (quoting S.Rep. No. 1333 at 6). Trademark protection helps both industry and consumers by protecting a source of important information, symbolic identification with a quality product, that consumers can use to aid their purchasing decisions. See J. Thomas McCarthy, Trademarks and Unfair Competition § 2.1 at 44-49 (2d ed. 1984).

Under Section 43(a) of the Lan-ham Act, 15 U.S.C. § 1125(a), a term is protectable as a trademark only if “the public recognizes it as identifying the claimant’s ‘goods or services and distinguishing them from those of others.’ ” Canfield, 808 F.2d at 296 (quoting McCarthy, § 15.1 at 657). Courts have adopted four categories of distinctiveness to aid in evaluations of a term’s inherent distinctiveness. This Court has described those categories as:

arbitrary (or fanciful) terms, which bear “no logical or suggestive relation to the actual characteristics of the goods;” suggestive terms which suggest rather than describe the characteristics of the goods; descriptive terms, which describe a characteristic or ingredient of the article to which it refers, and generic terms, which function as the common descriptive name of a product class.

Canfield, 808 F.2d at 296 (emphasis added) (quotations omitted). 2 If a term is arbitrary or suggestive, courts will treat it as distinctive and automatically qualify it for trademark protection. Id. at 297. If a term is descriptive, trademark protection will exist only if a claimant proves that the term conveys to consumers a secondary meaning of association with the claimant. Id. Finally, if a term is generic, courts are unwilling to afford it trademark protection. “Courts refuse to protect a generic term because competitors need it more to describe their goods than the claimed mark-holder needs it to distinguish its goods from others.” Id. at 304.

The district court held that the term “air door” was generic in 1962. In order thoroughly to evaluate the district court’s use of legal criteria in the present case, we must determine both what test of genericness should be applied in cases such as this and what evidence will be helpful in determining genericness under our articulated test.

*980 In making our choice of which test to apply, we find that both Congressional mandate and this court’s precedent direct that we employ the primary significance test. 3 Congress by statute has directed that the primary significance test is the criteria we should employ to determine ge-nericness. In 1984, Congress amended the Lanham Act to provide that “[t]he primary significance of the registered mark to the relevant public ... shall be the test for determining whether the registered mark has become the generic name of goods or services on or in connection with which it has been used.” Trademark Clarification Act of 1984, Pub.L. No. 98-620, Title I, § 102, 98 Stat. 3335 (1984) (codified at 15 U.S.C. § 1064). “The 1984 amendment makes the understanding of the ‘relevant public’ central to the genericness inquiry.” Magic Wand, Inc. v. RDB, Inc., 940 F.2d 638, 640 (Fed.Cir.1991). The Senate Report on the Trademark Clarification Act of 1984 emphasizes that the crucial subject of the Lanham Act genericness inquiry is the relevant consuming public. “The Committee concludes that it is necessary to clarify and reaffirm that the test of genericism is whether the relevant consuming public perceives a mark as an indication of source.” S.Rep. No. 627, 98th Cong., 2d Sess. 9-10 (1984). “In sum, the 1984 amendment makes the test for genericness the primary significance of the mark to the relevant public limited to the actual or potential purchasers of the goods or services.” Magic Wand, 940 F.2d at 641; see also Jerome Gilson, Trademark Protection and Practice § 2.02 at 2-29 (1992) (“It is the public understanding of the particular word which is critical [to genericness]”).

Turning next to our own precedent, this court has previously stated that “[t]he jurisprudence of genericness revolves around the primary significance test, which inquires whether the primary significance of a term in the minds of the consuming public is the product or the producer.” Canfield, 808 F.2d at 292-93. The test, as defined in Canfield, is not entirely appropriate here because in Canfield we were concerned with a product that had already been introduced into the marketplace. Here, we are using hindsight to inquire whether 30 years ago, when Mars was beginning to sell air curtains, the term “air door” was generic. The test as described in Canfield can, however, be molded to fit the present situation.

In utilizing the test in this case we cannot of course weigh whether the primary significance of “air door” in the minds of the consuming public was the product or the producer. Mars, the producer, was just entering the market and its product was not yet known to the public. The crucial inquiry at the introduction of the product must be instead whether the term “air door” in 1962 signified to the consuming public the product genus of “a barrier of moving air directed across an open doorway or other opening for temperature and insect control.” If it were demonstrated from a review of relevant publications, witness testimony, and expert analysis that the term “air door” had such a primary significance to consumers, then the term would be generic and Mars could not protect its use by trademark. See, e.g., In re Northland Aluminum Products, Inc., 777 F.2d 1556 (Fed.Cir.1985) (“bundt” was a common descriptive name of a ring style *981 cake and could not be registered as a trademark). If, on the other hand, the term did not have this significance to the relevant consumer, Mars would have had the right to protect its exclusive use of “air door.” See, e.g., In re Montrachet S.A., 878 F.2d 375 (Fed.Cir.1989) (“Montrachet” entitled to trade mark protection because the public did not perceive the term to be generic designation for a type of goat cheese).

In the present ease, the district court rejected using the primary significance or “buyer understanding test," see district court opinion at 7-8, and examined whether the “manufacturing industry” involved in the air curtain business commonly used the term “air door” to refer to its product before 1962. Id. at 9-13. We conclude that it was an error of law for the district court to do so.

In rejecting the primary significance test, the district court relied, not upon the pronouncements of Congress nor upon the previous determinations of this Court, but upon a recent case from the Western District of Missouri, First Nat’l Bank and Trust Co. v. First Nationwide Bank, 15 U.S.P.Q.2d 1457, 1459, 1990 WL 64768 (W.D.Mo.1990), which itself relies upon pronouncements of the Court of Appeals for the Seventh Circuit. E.g., Gimix, Inc. v. J.S. & A. Group, Inc., 699 F.2d 901, 905 (7th Cir.1983). The rationale of First National Bank and Gimix is that the primary significance test applies only to words, such as “Aspirin,” invented by the manufacturer that have allegedly become generic. See First Nat’l Bank, 15 U.S.P.Q.2d at 1460; Gimix, 699 F.2d at 905. The reasoning of the district coprt would have us apply a different test of genericness to an invented or new phrase than that applied to an invented or coined word.

We find this reasoning unpersuasive. First, the recent legislative and judicial pronouncements surrounding the 1984 amendments uniformly prescribe the primary significance test. Congress has not chosen to limit this test to apply only in cases of invented or coined words.

A distinction between “invented” words such as “aspirin” and new phrases such as “cash management account” is illogical. Combinations of old words to create new phrases can be just as specific product descriptive/nongeneric as new coined words. Moreover, when evaluating a phrase to determine whether it is generic, it must be examined as a whole. “Words which could not individually become a trademark may become one when taken together.” Blinded Veterans, 872 F.2d at 1041 (citation omitted). “The validity of a mark must be determined by looking at the mark as a whole. A composite mark should not be fragmented into its various pieces.” Id. (quoting McCarthy § 11:10 at 458). Even the Court of Appeals for the Seventh Circuit has recognized that “[dissecting marks often leads to error.” Liquid Controls, 802 F.2d at 934. 4 Just as the phrase "cash management account” was found to be non-generic by evaluating its meaning to the relevant consuming public, see In re Merrill Lynch, Pierce, Fenner, and Smith, 828 F.2d 1567, 1571 (Fed.Cir.1987), so too should “air door” be evaluated by examining its meaning to the relevant consuming public. 5 Even though the term “air door” may be the combination of two commonly used words, it is a phrase that *982 has some measure of inherent incongruity; a door is usually understood as being solid while air is not. 6 It is the relevant consumers, not the courts, who determine whether the term signifies the genus of “air barriers.”

The alternative to a test which focuses on consumer understanding is one based on the understanding either of judges or of the non-consuming public. But “[i]n determining whether a claimed trademark is generic or descriptive at the time of adoption, its meaning to a non-purchasing segment of the population is not important. The critical question is whether the mark is descriptive to the prospective purchasers of the product.” Blisscraft v. United Plastics Co., 294 F.2d 694, 699 (2d Cir.1961).

Finally, even were some distinction between invented words and invented phrases meaningful, limiting application of the primary significance test would contravene the purposes of trademark protection. Trademark protection is desirable because of the efficiencies and incentives produced by symbolic affiliation of producer and quality product. This principle relies in all cases upon a mark’s meaning to consumers. It is their affiliation of producer and product that will affect purchasing decisions. See McCarthy, § 2.1 at 44-49; Ralph H. Folsom & Larry L. Teply, Trademarked Generic Words, 89 Yale L.J. 1323, 1347 (1980). Regardless of whether a mark is an invented word or a developed phrase, “trademark protection benefits consumers by enabling them to select products on the basis of their origin ... [which] encourages sellers to create and maintain good will by marketing products of reliable quality that consumers associate with their mark.” American Cyanamid v. Connaught Labs., Inc., 800 F.2d 306, 308 (2d Cir.1986).

In order to accommodate Congress’ goals of promoting market efficiency and consumer understanding through trademark protection, the primary significance test must be utilized to determine a term’s ge-nericness. Only if a term’s meaning to the relevant consuming public is discovered can regulation of that term properly influence purchasing decisions and comport with relevant precedent.

Having established that the primary significance test must be used to evaluate the genericness of a term under the Lanham Act, it is important to discuss what kinds of evidence will be helpful in determining what a phrase means to the relevant consuming public. Generally, the relevant sources of information are numerous. “Evidence of the public’s understanding of the term [at issue] may be obtained from any competent source, such as purchaser testimony, consumer surveys, listings in dictionaries, trade journals, newspapers, and other publications.” In re Merrill Lynch, 828 F.2d at 1570. However, direct consumer evidence, e.g., consumer surveys and testimony is preferable to indirect forms of evidence. “Consumer surveys have become almost de rigueur in litigation over genericness. Judges are now used to survey evidence and often *983 expect to receive evidentiary assistance by surveys in resolving generic disputes.” McCarthy, § 12:2 at 529-30. 7

In the present case, the district court relied upon evidence found in trade journals and dictionaries. While both types of evidence may be helpful in Lanham Act cases, they are problematic. Articles in trade journals may indicate what a term means to the producer but not what it means to the relevant consuming public. As the court in In re Merrill Lynch, 828 F.2d at 1571, warned, “[t]he mixture of usages unearthed by the NEXIS computerized retrieval service does not show, by clear evidence, that the financial community views and uses the term CASH MANAGEMENT ACCOUNT as a generic, common descriptive term for the brokerage services to which Merrill Lynch first applied the term.” Moreover, dictionaries also “may not reflect word meaning among those persons who purchase the particular products involved.” Gilson, § 2.02 at 2-34. Additionally, there is likely to be a delay between a word’s acceptance into common usage and its entry in a dictionary. Dictionary entries also reflect lexicographical judgment and editing which may distort a word’s meaning or importance. “A Court accepting a dictionary entry at face value is in effect adopting the lexicographical judgment as its own, even though such a judgment might be based on printed matter which, if offered in evidence, would not be controlling.” Id. at 2-35.

Applying these principles to the present case, we conclude that the district court’s analysis was legally incorrect. First, the district court rejected without appropriate analysis Mars’ assertion that the “manufacturing industry” did not comprise the relevant consuming public for “air doors.” In doing so, it improperly rejected the primary significance test. While it is quite possible that members of the air curtain “manufacturing industry” may comprise a part of that consuming public, the evidence in the record indicates that others, including architects, construction companies and building managers, also comprise members of the relevant consuming public. See Appendix at 573. On remand, the district court should first determine the make-up of the relevant consuming public.

Second, the district court must determine what the term “air door” means to these consumers. In doing so, it should use direct consumer evidence if possible. Moreover, it should carefully scrutinize indirect evidence such as newspaper and trade journal articles and dictionary definitions. Even though the words “air” and “door” are generic enough in isolation, the phrase “air door” may very well have a unique, non-generic meaning to the relevant consuming population. See In re Merrill Lynch, 828 F.2d at 1570-71.

Because we reverse the district court’s analysis based upon its use of incorrect legal criteria, we need not address whether the district court improperly excluded evidence, as appellants allege. The district court excluded consumer survey evidence and expert testimony purporting to explain that evidence’s relevance to the time period at issue. This evidence was not considered by the district court in ruling on the motion for summary judgment in part because it was not properly submitted under the rules of procedure and in part because the court determined that it was of questionable relevance. On remand, the parties will have the opportunity to resubmit such evidence in a timely fashion. The district court should reevaluate its determination of relevance in accordance with this opinion. Moreover, even were the consumer survey evidence irrelevant to whether the term “air door” was generic in 1962, it would be relevant to whether the term is now gener *984 ic, an issue not currently before us but one that will be before the district court on remand.

IV.

For the foregoing reasons, we will reverse the district court’s summary judgment and injunction order and remand for proceedings consistent with this opinion.

. Both parties concede that a particular product may be described by more than one generic name.

. We are mindful that these categories are less than clearly defined and may be difficult to utilize. "These categories, like the tones in a spectrum, tend to blur at the edges and merge together. The labels are more advisory than definitional, more like guidelines than pigeon holes. Not surprisingly, they are somewhat difficult to articulate and apply.” Blinded Veterans Ass’n v. Blinded Am. Veterans Found., 872 F.2d 1035, 1039 (D.C.Cir.1989) (quoting Zatarains, Inc. v. Oak Grove Smokehouse, Inc., 698 F.2d 786, 790 (5th Cir.1983)).

. The primary significance test has its origin in the case of Bayer Co. v. United Drug Co., 272 F. 505, 509 (S.D.N.Y.1921), in which Judge Learned Hand held the term ASPIRIN to be generic. The court stated, "[t]he single question as I view it, in all these cases is merely one of fact: what do buyers understand by the word for whose use the parties are contending?” Id.

The primary significance test includes evaluating both the object and subject of identification. The object of identification is either a genus of goods or a specific good. "A generic term is one that is commonly used as the name of a kind of goods. Unlike a trademark, which identifies the source of a product, a generic term merely specifies the genus of which the particular product is a species." Liquid Controls, Corp. v. Liquid Control Corp., 802 F.2d 934, 936 (7th Cir.1986). The subject of identification is the relevant consuming public. "The critical issue in genericness cases is whether members of the relevant public primarily use or understand the term sought to be protected to refer to the genus of goods or services in question.” H. Marvin Ginn Corp. v. International Ass’n of Fire Chiefs, Inc., 782 F.2d 987, 989-90 (Fed.Cir.1986).

. Indeed, reliance upon either Liquid Controls or Gimix for the proposition that the primary significance test applies only to invented phrases is highly problematic. Liquid Controls never rejects the primary significance test and in fact includes analysis of what "liquid controls” means "in the minds of the public.” Liquid Controls, 802 F.2d at 938-39 n. 6. Gimix, in dictum, does specifically reject the primary significance test. However, the court there concluded that the phrase "Auto Page” was susceptible to many possible meanings and, thus, was not generic. Gimix, 699 F.2d at 906. Therefore, Gimix merely represents a pre-1984 amendment case where the primary significance test was purportedly rejected in dictum.

. Our analysis here is not inconsistent with our previous analysis in Canfield. In Canfield, we found that new phrase "chocolate fudge soda” was generic because it necessarily described a genus of products. 808 F.2d at 308. Unlike the term in Canfield which had no functional equivalents which could describe the genus at issue, the term “air door” has functional equivalents, such as "air curtain,” that describe the relevant genus.

. The district court reasoned that the phrase "air door" was not an invented phrase because it found that the evidence supported the conclusion that "the term ‘air door' was in use among members of the manufacturing industry well before defendant adopted the mark,” district court opinion at 9, and because "[t]he words ‘air’ and ‘door’ are such common terms in the English language that it seems unnecessary to seek their definitions and whether the composite of the two terms ‘air door’ connotes more than the sum of the parts in the context of this case is equally plain.” Id. at 10.

This analysis is flawed. First, Berner contends, not that "air door” has been historically used in the English language, but merely that it had become generic by 1962. Even if Mars didn’t invent the phrase, it had been invented within several years prior to 1962. Second, even under the rationale articulated in Gimix and First Nat’l Bank, a phrase comprised of generic words cannot itself be generic if it is susceptible to multiple meanings. See Gimix, 699 F.2d at 906. "Air door" could conceivably mean a door made of air, a doorway to air, a door made of a substance other than air but transparent or invisible, a doorway triggered by air, or a door that filters or impedes only air. Thus, we disagree that the term’s meaning is “equally plain.” Finally, the district court’s analytical framework is circular. It would require determining what test of genericness to apply based upon whether a term were "invented” and determining whether a term were invented based upon its "genericness.”

. The general value of consumer survey evidence is not disputed by our decision in Can-field. There, the court warned that consumer survey evidence that purported to link a term with a specific product rather than a genus of products could be misleading when the genus at issue contained only one specific product. 808 F.2d at 304. The court reasoned that consumer surveys linking "chocolate fudge” soda with Canfield were not dispositive evidence of non-genericness when Canfield had produced the only "chocolate fudge” soda on the market. Id. at 305. The unique problem of consumer survey evidence with regard to a single-product market or genus does not apply to other contexts such as the present case.