WOLLMAN, Circuit Judge.

The Clorox Company appeals from the district court's denial of its motion for preliminary injunctive relief brought within the context of its counterclaim against United Industries Corporation for false advertising under section 43(a) of the Lanham Act, 15 U.S.C. § 1125(a). We affirm.

I.

Clorox and United Industries are competing producers of roach bait insecticide products. Clorox manufactures and sells Combat, the top-selling brand of roach bait, while United Industries manufactures and sells the Maxattrax brand of roach bait, a small and relatively new participant in this market. United Industries initiated this action against Clorox, seeking a declaratory judgment that the packaging of its Maxattrax product, which predominantly asserts that it "Kills Roaches in 24 Hours," did not constitute false advertising or unfair competition under the Lanham Act. In response, Clorox moved to dismiss the complaint, contending that no actual case or controversy existed between the two parties regarding the packaging claims. Shortly thereafter, however, Clorox withdrew its motion to dismiss and filed an answer and counterclaim, which subsequently was amended. Clorox's amended counterclaim alleged, primarily, that a Maxattrax television commercial that United Industries had recently released for broadcast constituted false, deceptive, and misleading advertising in violation of section 43(a) of the Lanham Act, 15 U.S.C. § 1125(a)(1)(B).

The commercial at issue, entitled "Side by Side" by the advertising firm that produced it, depicts a split-screen view of two roach bait products on two kitchen countertops. The lighting is dark. On the left, one sees the Maxattrax box; on the right, a generic "Roach Bait" box that is vaguely similar to the packaging of the Combat brand sold by Clorox. An announcer asks the question: "Can you guess which bait kills roaches in 24 hours?" The lights then come up as the camera pans beyond the boxes to reveal a clean, calm, pristine kitchen, uninhabited by roaches, on the Maxattrax side. On the other side, the kitchen is in a chaotic state: cupboards and drawers are opening, items on the counter are turning over, paper towels are spinning off the dispenser, a spice rack is convulsing and losing its spices, all the apparent result of a major roach infestation. At the same time, the message "Based on lab tests" appears in small print at the bottom of the screen. The two roach bait boxes then reappear on the split-screen, and several computer-animated roaches on the "Roach Bait" side appear to kick over the generic box and dance gleefully upon it. The final visual is of the Maxattrax box only, over which the announcer concludes, "To kill roaches in 24 hours, it's hot-shot Maxattrax. Maxattrax, it's the no-wait roach bait." The final phrase is also displayed in print on the screen. The entire commercial runs fifteen seconds.

Clorox filed a motion for a preliminary injunction against this commercial. After expedited discovery and a two-day hearing, the district court denied the motion.

II.

In deciding a motion for a preliminary injunction, district courts are instructed to consider what have come to be known as the Dataphase factors:

(1) The probability of success on the merits; (2) The threat of irreparable harm to the movant; (3) The balance between this harm and the injury that granting the injunction will inflict on other interested parties; and (4) Whether the issuance of an injunction is in the public interest.

Sanborn Mfg. Co., Inc. v. Campbell Hausfeld/Scott Fetzer Co., 997 F.2d 484, 485-86 (8th Cir. 1993) (citing Dataphase Sys., Inc. v. C L Sys., Inc., 640 F.2d 109, 114 (8th Cir. 1981) (en banc)); see also Minnesota Mining and Mfg. Co. v. Rauh Rubber, Inc., 130 F.3d 1305, 1307 (8th Cir. 1997). No single factor in itself is dispositive; rather, each factor must be considered to determine whether the balance of equities weighs toward granting the injunction. See Sanborn, 997 F.2d at 486; Calvin Klein Cosmetics Corp. v. Lenox Lab., Inc., 815 F.2d 500, 503 (8th Cir. 1987).

We have noted that by enacting the Lanham Act, Congress apparently intended to encourage competitors to seek injunctions as a method of combating false advertising, and, in such cases that ultimately prove to have merit, injunctive relief is not to be issued reluctantly. See Black Hills Jewelry Mfg. Co. v. Gold Rush, Inc., 633 F.2d 746, 753 n. 7 (8th Cir. 1980). With regard to a preliminary injunction, however, the burden on the movant is heavy, in particular where, as here, "granting the preliminary injunction will give [the movant] substantially the relief it would obtain after a trial on the merits." Sanborn, 997 F.2d at 486 (quoting Dakota Indus., Inc. v. Ever Best Ltd., 944 F.2d 438, 440 (8th Cir. 1991)). Caution must therefore be exercised in a court's deliberation, and "the essential inquiry in weighing the propriety of issuing a preliminary injunction is whether the balance of other factors tips decidedly toward the movant and the movant has also raised questions so serious and difficult as to call for more deliberate investigation." General Mills, Inc. v. Kellogg Co., 824 F.2d 622, 624-25 (8th Cir. 1987).

A district court has broad discretion when ruling on requests for preliminary injunctions, and we will reverse only for clearly erroneous factual determinations, an error of law, or an abuse of that discretion. See Sanborn, 997 F.2d at 486 (citing Calvin Klein, 815 F.2d at 503). A district court's finding is "clearly erroneous" when, although evidence may exist to support it, upon review of the entire record we are left with the definite and firm conviction that error has occurred. See Prufrock Ltd., Inc. v. Lasater, 781 F.2d 129, 133 (8th Cir. 1986); Anderson v. City of Bessemer City, 470 U.S. 564, 573-74 (1985). This deferential standard of review "rests upon the unique opportunity afforded the trial court judge to evaluate the credibility of witnesses and to weigh the evidence." Inwood Lab., Inc. v. Ives Lab., Inc., 456 U.S. 844, 855 (1982).

A.

In our analysis of the Dataphase factors, we begin by assessing the probability of Clorox's ultimate success on the merits. At the early stage of a preliminary injunction motion, the speculative nature of this particular inquiry militates against any wooden or mathematical application of the test. See Calvin Klein, 815 F.2d at 503. Instead, "a court should flexibly weigh the case's particular circumstances to determine whether the balance of equities so favors the movant that justice requires the court to intervene to preserve the status quo until the merits are determined." Id. (quoting Dataphase, 640 F.2d at 113). Clorox asserts Lanham Act violations and seeks, primarily, permanent injunctive relief against the alleged false claims contained in the Maxattrax "Side by Side" commercial.

Section 43(a) of the Lanham Act

The Lanham Act was intended, in part, to protect persons engaged in commerce against false advertising and unfair competition. See Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763, 767-68 (1992) (quoting 15 U.S.C. § 1127); 3 J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition § 27:25 at 27-40 (West Group 1997). In particular, the Act prohibits commercial advertising or promotion that misrepresents the nature, characteristics, qualities, or geographic origin of the advertiser's or another person's goods, services, or commercial activities. See Rhone-Poulenc Rorer Pharm., Inc. v. Marion Merrell Dow, Inc., 93 F.3d 511, 514 (8th Cir. 1996) (citing 15 U.S.C. § 1125(a)(1)(B)); Sanborn, 997 F.2d at 486.

To establish a claim under the false or deceptive advertising prong of the Lanham Act, a plaintiff must prove: (1) a false statement of fact by the defendant in a commercial advertisement about its own or another's product; (2) the statement actually deceived or has the tendency to deceive a substantial segment of its audience; (3) the deception is material, in that it is likely to influence the purchasing decision; (4) the defendant caused its false statement to enter interstate commerce; and (5) the plaintiff has been or is likely to be injured as a result of the false statement, either by direct diversion of sales from itself to defendant or by a loss of goodwill associated with its products. See Southland Sod Farms v. Stover Seed Co., 108 F.3d 1134, 1139 (9th Cir. 1997); Johnson Johnson-Merck Consumer Pharm. Co. v. Rhone-Poulenc Rorer Pharm., Inc., 19 F.3d 125, 129 (3d Cir. 1994). In addition, to recover money damages under the Act, a "[p]laintiff must prove both actual damages and a causal link between defendant's violation and those damages." Rhone-Poulenc, 93 F.3d at 515.

The false statement necessary to establish a Lanham Act violation generally falls into one of two categories: (1) commercial claims that are literally false as a factual matter; and (2) claims that may be literally true or ambiguous but which implicitly convey a false impression, are misleading in context, or likely to deceive consumers. See Southland, 108 F.3d at 1139; National Basketball Ass'n v. Motorola, Inc., 105 F.3d 841, 855 (2d Cir. 1997); Abbott Lab. v. Mead Johnson Co., 971 F.2d 6, 13 (7th Cir. 1992). Many claims will actually fall into a third category, generally known as "puffery" or "puffing." Puffery is "exaggerated advertising, blustering, and boasting upon which no reasonable buyer would rely and is not actionable under § 43(a)." Southland, 108 F.3d at 1145; see also Castrol Inc. v. Pennzoil Co., 987 F.2d 939, 945 (3d Cir. 1993). Non-actionable puffery includes representations of product superiority that are vague or highly subjective. See Southland, 108 F.3d at 1145; Cook, Perkiss Liehe, Inc. v. Northern California Collection Serv., Inc., 911 F.2d 242, 246 (9th Cir. 1990) (advertising that merely states in general terms that one product is superior is not actionable). However, false descriptions of specific or absolute characteristics of a product and specific, measurable claims of product superiority based on product testing are not puffery and are actionable. See Southland, 108 F.3d at 1145; Castrol, 987 F.2d at 945.

1. Literally false claims

If a plaintiff proves that a challenged claim is literally false, a court may grant relief without considering whether the buying public was actually misled; actual consumer confusion need not be proved. See Rhone-Poulenc, 93 F.3d at 516; Johnson Johnson-Merck, 19 F.3d at 129; McNeil-P.C.C., Inc. v. Bristol-Myers Squibb Co., 938 F.2d 1544, 1549 (2d Cir. 1991) (where advertisement is shown to be literally false, court may enjoin it without reference to its impact on consumers). In assessing whether an advertisement is literally false, a court must analyze the message conveyed within its full context. See Rhone-Poulenc, 93 F.3d at 516; Southland, 108 F.3d at 1139. In some circumstances, even a visual image, or a visual image combined with an audio component, may be literally false:

We find, therefore, that the squeezing-pouring sequence in the Jenner commercial is false on its face. The visual component of the ad makes an explicit representation that Premium Pack is produced by squeezing oranges and pouring the freshly-squeezed juice directly into the carton. This is not a true representation of how the product is prepared. Premium Pack juice is heated and sometimes frozen prior to packaging.

Coca-Cola Co. v. Tropicana Products, Inc., 690 F.2d 312, 318 (2d Cir. 1982); see also Rhone-Poulenc, 93 F.3d at 516 (drug manufacturer's advertisements featuring images such as two similar gasoline pumps or airline tickets with dramatically different prices, accompanied by slogan, "Which one would you choose?" was literally false message that competing drugs could be indiscriminately substituted). The greater the degree to which a message relies upon the viewer or consumer to integrate its components and draw the apparent conclusion, however, the less likely it is that a finding of literal falsity will be supported. Commercial claims that are implicit, attenuated, or merely suggestive usually cannot fairly be characterized as literally false.

The district court determined that the Maxattrax commercial conveyed an explicit message that the product killed roaches in 24 hours and found that this message was literally true. The court concluded that scientific testing performed both by United Industries and Clorox sufficiently demonstrated that Maxattrax, which contains the fast-acting nerve toxin known as chlorpyrifos or Dursban, will actually kill a roach within 24 hours of its coming into contact with the product. In response, Clorox argues that the district court erroneously "ignored the explicit visual statements in United's advertising that, as a matter of law, combine with its express audio statements to determine its literal meaning." Brief for Appellant at 16. Clorox contends that the Maxattrax commercial conveyed three additional explicit messages that are literally false: (1) that Maxattrax controls roach infestations in consumers' homes within 24 hours; (2) that Combat and other roach baits are entirely ineffective in consumers' homes within 24 hours; and (3) that Maxattrax provides superior performance in consumers' homes in comparison to Combat and other roach baits.

Our review of the record satisfies us that the district court's determination that the commercial was literally true is not clearly erroneous. The court was clearly correct in its assessment that the audio and print components of the advertisement are literally true. The scientific evidence and expert testimony contained in the record satisfactorily established that Maxattrax roach bait "kills roaches in 24 hours." Clorox protests that this statement is literally true only in circumstances where a particular roach actually comes into the contact with the product. This complaint rings hollow. The requirement that roaches must come into contact with the poison for it to be effective is the central premise of the roach bait line of products. We will not presume the average consumer to be incapable of comprehending the essential nature of a roach trap.

Similarly, we conclude that the district court did not err in determining that the Maxattrax commercial did not convey explicit visual messages that were literally false. The depiction of a Maxattrax box in a pristine, roach-free kitchen, coupled with the depiction of a kitchen in disarray in which animated roaches happily dance about on a generic roach trap, is not sufficient, in our view, to constitute literal falsity in the manner in which it was presented. When the context is considered as a whole, moreover, the audio component of the advertisement, emphasizing only the 24-hour time frame and quick roach kill with no mention of complete infestation control, fosters ambiguity regarding the intended message and renders the commercial much more susceptible to differing, plausible interpretations. Thus, in our view, the district court's finding that the commercial did not explicitly convey a literally false message that Maxattrax will completely control a home roach infestation within 24 hours is not clearly erroneous.

Clorox also contends that the commercial conveys an explicit message of comparative superiority that is literally false. We have recently distinguished between two types of comparative advertising claims brought under the Lanham Act: (1) "my product is better than yours" and (2) "tests prove that my product is better than yours." Rhone-Poulenc, 93 F.3d at 514 (emphasis in original). When challenging a claim of superiority that does not make express reference to testing, a plaintiff must prove that the defendant's claim of superiority is actually false, not simply unproven or unsubstantiated. See id. Under a "tests prove" claim, in which a defendant has buttressed a claim of superiority by attributing it to the results of scientific testing, a plaintiff must prove only "that the tests [relied upon] were not sufficiently reliable to permit one to conclude with reasonable certainty that they established the proposition for which they were cited." Id. at 514-15 (quoting Castrol, Inc. v. Quaker State Corp., 977 F.2d 57, 62-63 (2d Cir. 1992)). However, "[t]o ensure vigorous competition and to protect legitimate commercial speech, courts applying this standard should give advertisers a fair amount of leeway, at least in the absence of a clear intent to deceive or substantial consumer confusion." Rhone-Poulenc, 93 F.3d at 515.

The Maxattrax commercial indicates in small print at the bottom of the screen that its implied answer to the posed question, "Can you guess which bait kills roaches in 24 hours?" is, "Based on lab tests." In order for this claim to be considered literally false, then, Clorox must establish that the tests to which the commercial referred were not sufficiently reliable to support its claims with reasonable certainty. See id. at 514-15. The district court determined that the scientific research provided by United Industries was reliable and supported the commercial's claims. We agree with this conclusion. Laboratory testing indicates that the toxin contained in Maxattrax kills within 24 hours those roaches that come into contact with it. Some other roach bait products will not kill a roach within that interval and, in fact, are not even intended to do so.

Any additional messages in the Maxattrax commercial perceived by Clorox, visual or otherwise, are not sufficiently explicit or unambiguous so as to constitute specific false claims of a literal nature. Thus, we cannot say that the court committed clear error in its determinations regarding the scope of the commercial's explicit claims of superiority (that it kills roaches within 24 of hours and that a generic competitor does not), or in finding that claim to be literally true. See L F Products, a Div. of Sterling Winthrop, Inc. v. Proctor Gamble Co., 45 F.3d 709, 712 (2d Cir. 1995) (district court's determination with respect to facial falsity was not clearly erroneous).

2. Implicitly false or misleading claims

Statements that are literally true or ambiguous but which nevertheless have a tendency to mislead or deceive the consumer are actionable under the Lanham Act. See Southland, 108 F.3d at 1140; Sandoz Pharm. Corp. v. Richardson-Vicks, Inc., 902 F.2d 222, 228-29 (3d Cir. 1990); American Home Products Corp. v. Johnson Johnson, 577 F.2d 160, 165 (2d Cir. 1978). Where a commercial claim is not literally false but is misleading in context, proof that the advertising actually conveyed the implied message and thereby deceived a significant potion of the recipients becomes critical. See William H. Morris Co. v. Group W, Inc., 66 F.3d 255, 258 (9th Cir. 1995) (per curiam); Johnson Johnson-Merck Consumer Pharm. Co. v. Smithkline Beecham Corp., 960 F.2d 294, 297-98 (2d Cir. 1992).

If a plaintiff does not prove the claim to be literally false, he must prove that it is

deceptive or misleading, which depends on the message that is conveyed to consumers. Public reaction is the measure of a commercial's impact. As the district court noted, the success of the claim usually turns on the persuasiveness of a consumer survey.

Johnson Johnson-Merck, 19 F.3d at 129-30 (internal citations omitted).

In affirming a jury verdict awarding damages under a Lanham Act claim, we recently held that a manufacturer was not required to provide consumer surveys or reaction tests in order to prove entitlement to damages in a false comparative advertising action against its competitor where the jury found that the competitor had violated the Lanham Act willfully and in bad faith. See Porous Media Corp. v. Pall Corp., 110 F.3d 1329, 1337 (8th Cir. 1997). Where, as here, there has been no finding of a willful violation or an intent to deceive, evidence of consumer impact is essential. See William H. Morris, 66 F.3d at 258-59; American Tel. and Tel. Co. v. Winback and Conserve Program, Inc., 42 F.3d 1421, 1443 (3d Cir. 1994); Johnson Johnson-Merck, 19 F.3d at 129-30; Abbott Lab., 971 F.2d at 14; Smithkline Beecham, 960 F.2d at 297-98; Coca-Cola, 690 F.2d at 317. Therefore, unless a commercial claim is literally false, or a trier of fact has determined that a competitor acted willfully with intent to deceive or in bad faith, a party seeking relief under this section of the Lanham Act bears the ultimate burden of proving actual deception by using reliable consumer or market research. See Smithkline Beecham, 960 F.2d at 297 ("It is not for the judge to determine, based solely upon his or her intuitive reaction, whether the advertisement is deceptive."); ATT, 42 F.3d at 1443 (quoting Sandoz, 902 F.2d at 228-29) ("[I]t cannot obtain relief by arguing how consumers could react; it must show how consumers actually do react.").

At the preliminary injunction stage, however, full-blown consumer surveys or market research are not an absolute prerequisite, and expert testimony or other evidence may at times be sufficient to obtain preliminary injunctive relief in cases involving implicitly false or misleading claims. See Abbott, 971 F.2d at 15; 3 McCarthy § 27:55 at 27-81 ("However, on a motion for a preliminary injunction, a survey is not always necessary and it is sufficient if plaintiff introduces expert testimony or any other evidence showing that a significant number of consumers received the claimed message from the advertisement.").

Clorox contends that when one assesses the comparative visuals and implicit messages in the commercial, a consumer might be misled to construe them as a claim that Maxattrax will completely control an infestation by killing all of the roaches in one's home within 24 hours, while its competitors will fail to do the same. In fact, Maxattrax will kill only those roaches which come into contact with the product; actual control of a roach problem may take several weeks. Whether one accepts the district court's more literal interpretation of the commercial's message or Clorox's proposed construction, however, is highly dependent upon context and inference, and Clorox's view is unsupported at this point by expert testimony, surveys, or consumer reaction evidence of any kind. It is, in other words, a classic question of fact, the resolution of which we will not disturb absent a showing of clear error by the district court. Clorox has not made such a showing.

In sum, then, the district court did not err in concluding that Clorox had not shown a likelihood of success on the merits of the claim.

B.

The remaining Dataphase factors do not tip the balance of equities decidedly in favor of Clorox. We have stated that the failure to demonstrate the threat of irreparable harm is, by itself, a sufficient ground upon which to deny a preliminary injunction. See Adam-Mellang v. Apartment Search, Inc., 96 F.3d 297, 299 (8th Cir. 1996). When injunctive relief is sought under the Lanham Act, the finding of a tendency to deceive satisfies the requisite showing of irreparable harm. See Black Hills Jewelry, 633 F.2d at 753 ("To obtain an injunction under section 43(a) appellees need only show that the falsities complained of had a tendency to deceive."); McNeilab, Inc. v. American Home Products Corp., 848 F.2d 34, 38 (2d Cir. 1988) (where challenged advertisement directly, but falsely, proclaims superiority of defendant's product over plaintiff's, irreparable harm may be presumed). Absent such a showing, however, irreparable harm cannot be presumed where, as here, plaintiff has not established any prospect of success upon the merits. See Sanborn, 997 F.2d at 489; Johnson Johnson v. Carter-Wallace, Inc., 631 F.2d 186, 192 (2d Cir. 1980) ("While proof of actual diversion of sales is not required for a § 43(a) injunction to issue, proof that the advertising complained of is in fact false is essential."). Clorox has not otherwise sufficiently demonstrated the threat of irreparable injury so as to tip this factor in its favor.

The district court did not make an explicit finding concerning the balance of harm to Clorox stemming from the commercial and the injury to United Industries that would result from an injunction. In light of the district court's conclusion that Clorox had failed to demonstrate a probability of ultimate success, the possibility that it will suffer any harm from the continuing airing of the commercial is highly speculative and therefore does not serve to tip the balance of equities in Clorox's favor. See generally Sanborn, 997 F.2d at 489-90.

Finally, "[a]lthough the public interest favors enjoining false statements," id. at 490, absent a more substantial showing that Clorox has a viable claim, this factor likewise does not tilt the equities toward granting preliminary injunctive relief. See id. Therefore, because we believe that it committed no clear error in its factual findings or legal conclusions, we conclude that the district court did not abuse its broad discretion in denying Clorox's motion for preliminary injunction.

III.

Additionally, Clorox challenges various statements made by United Industries regarding the alleged effectiveness and superiority of Maxattrax in its promotional materials distributed to retailers, some of which specifically referred to Combat. Such materials may properly be the basis of a claim under section 43(a) of the Lanham Act. See, e.g., Seven-Up Co. v. Coca-Cola Co., 86 F.3d 1379, 1382-83 (5th Cir. 1996). The district court held that Clorox was not entitled to injunctive relief regarding these materials because it had never alleged that such materials were violative of the Lanham Act. See Findings of Fact, Conclusions of Law and Order at 7-8. Clorox sought post-hearing leave to amend, nunc pro tunc, its counterclaim and motion for preliminary injunction to include claims based upon allegedly false statements contained within the promotional literature, which the district court apparently denied.

Although leave to amend is to be freely granted, a district court's denial of such leave is reviewed for abuse of discretion. See Springdale Educ. Ass'n v. Springdale Sch. Dist., 133 F.3d 649, 653 (8th Cir. 1998); In re NationsMart Corp. Sec. Litig., 130 F.3d 309, 322 (8th Cir. 1997). We find no abuse of discretion in the district court's denial of leave to amend. Whether these materials should be considered during a trial on the merits for permanent injunctive relief is a matter to be addressed by the district court.

The order denying preliminary injunctive relief is affirmed.