Church Dwight v. S.C. Johnson Son

Full title: CHURCH DWIGHT CO., INC., Plaintiff, v. S.C. JOHNSON SON, INC., Defendant

Court: United States District Court, D. New Jersey

Date published: Nov 23, 1994

873 F. Supp. 893 (D.N.J. 1994)

Opinion

Civ. No. 94-2829.

November 23, 1994.

Thomas C. Morrison and Frederick B. Warder III of Paterson, Belknap, Webb Tyler, New York City, for plaintiff Church Dwight Co., Inc.

Lawrence I. Weinstein and Robert L. Sherman of Paul, Hastings, Janofsky Walker, New York City, for defendant S.C. Johnson Son, Inc.

BROWN, District Judge

This matter comes before the Court on plaintiff Church Dwight Co. Inc.'s motion for permanent injunctive relief. Specifically, plaintiff has applied to this Court for an order permanently enjoining defendant S.C. Johnson Son, Inc. from advertising that its carpet deodorizing products are "five times better" than baking soda. For the reasons set forth below, which constitutes this Court's findings of fact and conclusions of law pursuant to FED.R.CIV.P. 52(a), the Court will grant plaintiff's motion for permanent injunctive relief.

I. BACKGROUND FINDINGS OF FACT AND PROCEDURAL HISTORY

The factual background of this action, as well as some of the parties' contentions, are set forth below. The hearing conducted before the Court consisted generally of cross-examination and re-direct examination of the witnesses presented, as all direct testimony and evidence was presented through affidavits and trial exhibits, except for certain limited direct examination requested by the parties.

A. COMMON HOUSEHOLD ODORS AND METHODS OF ODOR ELIMINATION

The pungent odors of cat urine and cigarette smoke permeate this litigation, as the parties to this action, Church Dwight, Co. Inc. (hereinafter "Church Dwight") and S.C. Johnson Sons, Inc. (hereinafter "S.C. Johnson") market various deodorizing products targeted at eliminating these unpleasantries. To fully comprehend this lawsuit, however, one must first understand the concept of the term "odor."

The Third Edition of the American Heritage Dictionary defines "odor" as "(1) the property or quality of a thing that affects, stimulates or is perceived by the sense of smell; (2) a sensation, stimulation, or perception of the sense of smell." See Pl.Direct Evidence, Testimony of Dr. Amos Turk at ¶ 7. As plaintiff's expert Dr. Amos Turk testified,

odors are not substances; they are properties or perceptions. To put this in more concrete terms, cat urine and cigarette smoke are not odors. They are substances that create odors by stimulating a human sensory perception, and thus may be called "odorants." Because they generate unpleasant odors, they are often referred to as "malodorants," and the odors they generate as "malodors."

Id. Virtually all malodors are volatile organic compounds (hereinafter "VOCs"). See Affidavit of Dr. Peter E. Nott at ¶ 9. Defendant's expert, Dr. Peter E. Nott explained that "[a] substance is referred to as being `volatile' if some of its molecules exist in the air, in the gaseous or vapor state. If an odor is not volatile, people would not perceive it because the air would not carry it to the nose." Id. Dr. Nott further stated that:

It is uniformly accepted as scientific fact that the perceived strength of an odor is a function of the concentration of the odor molecules, or VOCs, in the air. If you reduce the concentration of the odor molecule, you will reduce the perception of the odor. Virtually all odors, and certainly pet and smoke odors, are made up primarily of VOCs. Thus, by accurately measuring the amount of odor molecules, or VOCs, before and after treatment with a deodorizer, one is able to develop a good picture of the efficacy of the product.

Id. at ¶ 11.

There are three basic methods used to eliminate malodorant VOCs: (1) absorption; (2) adsorption; and (3) neutralization. Id. at ¶¶ 8-11. Absorption involves a material binding or adhering to an internal structure of another material, similar to a sponge absorbing a liquid. See Affidavit of Dr. Daniel M. Ennis at ¶ 17; see also Affidavit of Dr. Peter E. Nott at ¶ 20. Adsorption "refers to a material binding on an external surface of another material," thereby trapping the odor molecule to the surface area of the adsorbing material. See Affidavit of Dr. Daniel M. Ennis at ¶ 17; see also Affidavit of Dr. Peter E. Nott at ¶ 20. The process of adsorption is a function of the adsorbing product's surface area. See Affidavit of Dr. Peter E. Nott at ¶ 22. Lastly, neutralization is a chemical process by which an acid and a base react to form a salt and water. See Affidavit of Dr. Daniel M. Ennis at ¶ 17; see also Affidavit of Dr. Peter E. Nott at ¶ 20. Thus, to eliminate common household odors, such as cat urine and cigarette smoke, the deodorizing products marketed by plaintiff employs a combination of adsorption and neutralization, while defendant's products use a combination of all three odor eliminating techniques. See infra discussion at Section B[1-2].

B. THE CARPET DEODORIZING PRODUCTS, THEIR CHEMICAL COMPOSITION AND THEIR METHODS OF ODOR ELIMINATION

1. Church Dwight's Arm Hammer Carpet Deodorizer

Plaintiff Church Dwight manufactures a line of carpet and room deodorizers under the Arm Hammer name. See Pl.Direct Evidence, Testimony of James E. Barch at ¶ 3. Although the Arm Hammer carpet and room deodorizers are available in a variety of scents, the active deodorizing agent is sodium bicarbonate, commonly known as baking soda. Id. at ¶¶ 4-5. Baking soda's deodorizing capacity is twofold: (1) neutralization and (2) adsorption. See Affidavit of Dr. Peter E. Nott at ¶¶ 21.

Baking soda — a weak base — chemically reacts with acidic odorants to form an odorless salt and water. See Pl.Direct Evidence, Testimony of Dr. Amos Turk at ¶ 9. Moreover, baking soda can also eliminate odors by adsorption, i.e., odorous molecules bind to the external surface of baking soda. See Affidavit of Dr. Peter E. Nott at ¶ 22.¹ "Because of the physical structure of baking soda, it has a relatively small surface area to effect adsorption." Id. In regards to pet urine and smoke odor, which consist of primarily of neutral and weak basic VOCs, baking soda's primary deodorizing capability is attributable to adsorption rather than neutralization. Id. at ¶ 21. In addition to the baking soda, Church Dwight's products also contain a mild fragrance which is used to "mask" the malodor. See Pl.Direct Evidence, Testimony of James E. Barch at ¶ 5.

1 Baking soda cannot absorb malodorous molecules because its chemical structure is not sufficiently porous. Id. at ¶ 21.

2. S.C. Johnson's Glade Carpet Deodorizers: Regular Glade and Wet'n Dry Glade Formulations

Beginning in 1992, defendant S.C. Johnson attempted to find a new ingredient "for [its] Glade products that would have a vastly larger surface area than baking soda," to absorb and adsorb malodorous VOCs. See Affidavit of Dr. Peter E. Nott at ¶ 23; see also Affidavit of Catherine Clemency at ¶¶ 9-10. After extensive experimentation, defendant S.C. Johnson isolated a particular structure of aluminum silicate known as zeolite, which eliminates malodorous VOCs through adsorption and absorption. See Affidavit of Dr. Peter E. Nott at ¶ 23. S.C. Johnson ultimately selected a particular zeolite that was a selective absorber "which work[ed] effectively without absorbing [S.C. Johnson's] product fragrances." Id. at ¶ 24. Thereafter, defendant S.C. Johnson conducted a test employing a Quantachrome Autoscan Mercury Porosimeter to compare the total surface area of zeolite with baking soda. Id. at ¶ 25. The test revealed that zeolite had 5.6 times the surface area of baking soda. Id. S.C. Johnson began using zeolites in its regular Glade formulation and its Wet'n Dry Glade formulation. See Affidavit of Catherine Clemency at ¶ 10.

Defendant S.C. Johnson's new regular Glade formulation is composed of 79% sodium sulfate, 17% baking soda, 3% aluminum silicate, and 1% fragrance. See Pl.Trial Ex. 21 at 1-2. Zeolites represent only 3% of the newly formulated regular Glade. Id. Defendant S.C. Johnson's new Wet'n Dry Glade formulation is composed of 94% baking soda, 4% sodium sulfate, 1% aluminum silicate, and 1% fragrance. Id. Zeolites represent only 1% of the new Glade Wet'n Dry formula. Id. Both of these products use a combination of adsorption, absorption and neutralization to eliminate malodorous VOCs. See Affidavit of Dr. Peter E. Nott at ¶ 23.

C. S.C. JOHNSON'S "ELECTRONIC NOSE" TESTS

In 1993, scientists at S.C. Johnson began conducting tests with a Photoionization Analyzer, which can accurately quantify the amount of VOCs in a controlled chamber. See Affidavit of Dr. Peter E. Nott at ¶ 12. The S.C. Johnson scientists referred to the Photoionization Analyzer as the "Electronic Nose." Id. The Electronic Nose is a highly sophisticated piece of equipment which detects VOCs through the process of photoionization. Id. at ¶ 13. Photoionization is a process by which an ultraviolet lamp emits a selected energy beam which, in turn, is absorbed by the odor molecules in the controlled chamber. Id. The odor molecules are thereby ionized, and the Photoionization Analyzer measures the amount of ionized odor molecules present in the controlled chamber. Id. In the present studies, the Photoionization Analyzer was calibrated to take six readings every five minutes for a period of twenty minutes. Id. at ¶ 28.

Defendant S.C. Johnson selected an ultraviolet lamp with energy of 10.2 electron volts because this voltage was "able to detect most classes of organic compounds, such as acids, esters. aldehydes, ketones, amines, polycyclic aromatic hydrocarbons, heterocyclic compounds and other odorous classes of compounds. These classes of compounds include chemicals found in such malodors as cat urine and cigarette smoke." Id. The S.C. Johnson scientists then conducted a battery of photoionization tests to quantify the efficacy of the newly formulated Glade products as compared to baking soda at reducing malodorous VOCs. Id. at ¶ 26. When conducting these tests, the scientists used 15 microliters of cat urine and 5 microliters of liquid cigarette smoke. Id. at ¶ 27. First, defendants compared the ability of 10 grams of Glade base, minus the fragrance, and 10 grams of baking soda "to reduce the VOCs produced by cat urine and cigarette smoke." Id. The scientists concluded that the Glade base reduced the VOCs emanating from the 15 microliters of cat urine at least 5 times greater than baking soda. Id. The scientists also concluded that the Glade base was able to reduce the VOCs emanating from the 5 microliters of cigarette smoke at least 5 times greater than baking soda. Id.

Next, the scientists applied 15 microliters directly to the Glade base and the baking soda using a microsyringe injected through a port in the controlled chamber. Id. at ¶ 30. The scientists concluded that "[t]he ratio of VOC reduction in Glade's favor was not merely 5 to 1 but infinite, as the baking soda actually increased the VOCs of cat urine, while Glade base reduced the VOCs." Id. To confirm these findings, the S.C. Johnson scientists utilized a Gas Chromatograph to evaluate the effect of baking soda on cat urine. Id. at ¶ 31. The Gas Chromatograph results also demonstrated that baking soda actually increased the VOCs produced by cat urine. Id.

The scientists also conducted a test in which they compared the pet urine and smoke odor reduction capacity of 50 grams of baking soda and 10 grams of Glade base. Id. at ¶ 32. "In these tests, which were repeated three times, Glade base reduced VOC concentrations from both smoke and cat urine greater than five times as much [as] baking soda. Indeed, the baking soda again actually increased the VOCs in the cat urine test." Id. The scientists also conducted experiments on actual carpet swatches using both the base ingredients and the actual carpet deodorizing products. Id. at ¶ 33-35. These tests also revealed that the Glade product was able to reduce the VOCs emanating from the cigarette smoke and cat urine at least 5 times greater than baking soda. Id. In addition to the regular Glade formulation, the S.C. Johnson scientists conducted photoionization tests with cat urine regarding the Glade Wet'n Dry formulation with only 1% zeolites. Id. at ¶ 36. The scientists concluded that "this formulation was also effective in reducing cat urine VOCs," but did not quantify any ratio as compared to pure baking soda. Id.

D. S.C. JOHNSON'S "FIVE TIMES BETTER" ADVERTISING CAMPAIGN ²

2 Prior to this litigation, Church Dwight had its own advertising superiority claim in which it alleged that Arm Hammer Carpet and Room Deodorizer had 3 times more odor fighting ingredients than any other brand. See Cross-Examination of James E. Barch, dated October 4, 1994. When Church Dwight initially made this claim, its Arm Hammer product actually had 3 times more baking soda than any other brand. Id. The newly reformulated Glade Wet'n Dry product, however, was composed of 94% baking soda. See Pl. Trial Ex. 21 at 1-2. Thus, Church Dwight's "3 times more odor fighting ingredients" claim was no longer viable. See Cross-Examination of James E. Barch, dated October 4, 1994. During the pendency of this litigation, S.C. Johnson counter-claimed that Church Dwight's "3 times" superiority claim was literally false. See Answer and Counterclaims at ¶ 14. In light of the new Glade Wet'n Dry formulation, plaintiff Church Dwight has agreed to cease advertising that its product contained "3 times" more ingredients. See Cross-Examination of James E. Barch, dated October 4, 1994. Hence, defendant S.C. Johnson's counter-claim is moot.

After conducting these various laboratory tests, defendant S.C. Johnson began conducting consumer research to "help determine the best way to position [their] restaged product." See Affidavit of Catherine Clemency at ¶ 14. Defendant S.C. Johnson's marketing team determined that "the `five times better than baking soda' option would not only help inform consumers of our factual findings, but also help dispel the myth that baking soda was the most effective ingredient available at absorbing common household odors found in carpets." Id. Moreover, S.C. Johnson determined that the "`[f]ive times better' [claim] offered good results while requiring less advertising expenditure and ranked highest in purchase interest." Id. at ¶ 15.

In November 1993, Sarah S. DiVall, S.C. Johnson's Associate Research Services Manager, supervised a qualitative market research project concerning the "five times better claim." See Rebuttal Affidavit of Sarah S. DiVall at ¶ 2. During the market tests, ten focus groups composed of three consumers each were exposed to three possible television advertisements, including the commercial which was eventually aired. Id. Thereafter, an independent moderator questioned the focus groups regarding what the potential S.C. Johnson's television commercials had communicated to them. Id. Sarah DiVall observed the consumer groups through a two-way mirror. See Cross-Examination of Sarah S. DiVall dated October 5, 1994. During cross-examination, Ms. DiVall testified that consumers linked the "five times better claim" to how the product actually worked at eliminating odors and did not associate the claim with a five-fold reduction in odor molecules. Id. Ms. DiVall conceded, however, that consumers understood the "five times better" message to mean that S.C. Johnson's products had a five-fold capability of eliminating odors that people smell. Id.

Thus, in April 1994, S.C. Johnson began its "Five Times Better" advertising campaign on television, on its product cartons, and in various magazines. See Pl.Trial Exs. 5, 6, 8, and 9. Both the audio and visual elements of the television commercial claim that Glade regular formulation and Glade Wet'n Dry formulation perform "Five Times Better" than baking soda. See Pl.Trial Ex. 8. In the audio segment of the television commercial, S.C. Johnson makes the following claims: (1) "Glade . . . is now five times better than baking soda"; (2) "It's five times tougher"; (3) "absorbs tough carpet odors, five times better than baking soda"; (4) "It's five times fresher"; and (5) "It's five times better than baking soda." Id. Defendant S.C. Johnson reinforces its five-fold claim through the use of the following visual elements: (1) five men, sitting around a dining room table, smoking cigarettes in the third frame, followed by five empty chairs around the same dining room table in the fourth frame, and (2) five golden retriever puppies allegedly urinating on a living room carpet in the seventh frame, followed by three puppies in the eighth frame, only one puppy in the ninth frame, and no puppies remaining in the tenth frame in which a word overlay appears stating "Five Times Better Than Baking Soda." Id.

The "Five Times Better" claim also appeared on the various Glade packages. Id. at Exs. 5-6. Specifically, in large, red and black lettering on a yellow border, on the top front of the package appeared: ABSORBS ODORS 5 TIMES BETTER than BAKING SODA." Id. Additionally, it is worth noting that the same message appeared on the back of the package. Id. Also, in April, 1994, defendant S.C. Johnson began advertising its newly reformulated products in various women's magazines. Id. at Ex. 9. In the print advertisement, S.C. Johnson claims that its new products "absorb five times better than baking soda. That's five times the absorbing power to keep your home smelling fresh and clean." Id. Moreover, at the bottom of the print advertisement, in large, bold lettering, S.C. Johnson asserts: It's five times better." Id.

E. VARIOUS SENSORY TESTS PERFORMED

1. Plaintiff Church Dwight's First Internal Sensory Study

After learning of S.C. Johnson's "Five Times Better" advertising campaign, plaintiff Church Dwight conducted its own in-house sensory study to compare the efficacy of the newly formulated Glade products with baking soda. See Pl.Direct Evidence, Testimony of Raymond S. Brown at ¶ 13. Specifically, Church Dwight conducted its sensory tests on the following products: "Glade Potpourri Carpet and Room Deodorizer Spring Orchard; Glade Potpourri Wet'n Dry Formula Carpet and Room Deodorizer Fresh Scent; Arm Hammer Potpourri Carpet Deodorizer Country Scent; Arm Hammer Potpourri Carpet Deodorizer Country Scent; and Arm Hammer Baking Soda." Id. at ¶ 14. The odor panelists used in this first study "were drawn from a pool of Church Dwight employees who have been trained and are experienced in rating the intensity of odors. . . ." Id. at ¶ 16. Because some of the test products contained fragrance, Church Dwight attempted to design its sensory test to best control the fragrance effect. Id. at ¶ 15. It "did this by conducting the test over an extended period to allow the fragrance to dissipate and by directing [the] odor panelists to rate only the intensity of the malodor itself and to ignore the fragrance." Id.

To conduct the sensory test, Church Dwight employed carpet swatches that had been exposed to cigarette smoke, cat urine, and mildew malodors. Id. at ¶ 18. "Following odorization, the test products were applied to the carpet swatches in an amount reflecting average consumer use, and were then vacuumed up. Some of the malodorized swatches were not treated with any test product, and were instead used by the panelists as malodor references." Id. at ¶ 19. After application of the test products, the panelists evaluated the carpet swatches in paired comparisons at 2, 24, 48, and 72 hours. Id. at ¶ 20-21. The panelists were instructed that the first swatch they "were exposed to should be considered to have a malodor intensity rating of 10. The panelists were then instructed to rate the malodor intensity of the second swatch in relation to the first." Id. at ¶ 21. According to Church Dwight, these sensory tests revealed that there was "little, if any, difference in the odor elimination properties of the Glade products when compared to baking soda. When compared to the Arm Hammer carpet deodorizer, the Glade product performed at parity." Id. at ¶ 23; see also Pl.Trial Ex. 21.

2. The TRC Sensory Study

Following the first internal sensory study, Church Dwight "determined that it would be appropriate to engage an outside independent testing firm to conduct independent sensory tests." See Pl.Direct Evidence, Testimony of Raymond S. Brown at ¶ 24. Thus, Church Dwight retained Dr. Amos Turk and TRC Environmental Corp. (hereinafter "TRC") to conduct additional sensory tests. Id. at ¶¶ 24-26. The TRC study tested the following products: Arm Hammer Baking Soda; Arm Hammer Carpet Deodorizer — Original Potpourri; Glade Potpourri Carpet Room Deodorizer — Country Garden; and, Glade Potpourri Carpet and Room Deodorizer — Wet'n Dry Formula Fresh Scent. See Pl.Trial Ex. 22 at 1. The TRC study was conducted as follows:

1 square foot segments of plush nylon carpet were odorized with either cat urine or cigarette smoke, treated with test product, vacuumed and then evaluated by TRC's trained odor panel 2, 24, 48 and 72 hours after product treatment for perceived malodor level. The panelists were asked to rank the intensity of the malodor only (i.e. ignore the fragrance of the deodorizer) on a scale of 0 to 20, with 0 as the no odor level, and 20 as the extremely high odor level.

Id. The TRC study revealed that the Glade products did not reduce the odor intensity of cat urine and cigarette smoke by a five-fold decrease. Id. at 4. Rather, the TRC panelists concluded that, at 24 hours, the Glade products produced only a 2.16 fold decrease in cat urine malodor as compared to baking soda. Id. Ultimately, the TRC test demonstrated that "[t]here was no significant difference in the efficacy of the four carpet deodorizers in their ability to decrease the intensity of malodors over time." Id. at 6.

3. Plaintiff Church Dwight's Second Internal Sensory Study

Because both Church Dwight's first internal sensory study and TRC's study used fragranced products, Church Dwight endeavored to conduct a sensory study with unfragranced products. See Pl.Direct Evidence, Testimony of Raymond S. Brown at ¶ 27. Thus, after obtaining the unfragranced Glade product base for the regular and Wet'n Dry formulations from S.C. Johnson, Church Dwight conducted a second sensory test. Id. at ¶ 28. Church Dwight employed the same procedures in the second sensory test as were used in its first sensory test. Id. at ¶ 29. Ultimately, Church Dwight's second sensory study revealed that

once again, . . . glade products do not "absorb odors five times better than baking soda." At none of the evaluation periods did the ratios even come close to 5:1. In fact, the performance ratios are even closer in this test than in the previous tests: the largest ratio of glade's superiority to baking soda is 1.39:1, and most are closer to parity. Baking soda is in fact rated superior to Glade at a number of the evaluation periods.

Id. at ¶ 30.³

3 Plaintiff Church Dwight also commissioned Arthur D. Little Inc. (hereinafter "Little") to conduct a study concerning defendant S C. Johnson's "five times better" claim. See Defendant's Trial Exhibits, Volume II, Ex. 50. Specifically, Little conducted gravimetric tests and breakthrough tests on various carpet deodorizers and baking soda. Id. at 1. 3. "The design and details of each test resulted from a series of discussions with Dr. Amos Turk, [Church Dwight's] technical advisor for these studies. Id. at 3. On August 25, 1994, while being deposed, Dr. Amos Turk stated that because of flawed scientific protocol, the Little study was "useless." See Defendant's Trial Exhibits, Volume I, Deposition of Dr. Amos Turk at 397-98 (emphasis added). In fact, Church Dwight's counsel conceded at that deposition that it did not intend to rely on the Little study at trial. Id. at 336-39.

Moreover, in his initial affidavit dated September 21, 1994, in support of plaintiff's application for a preliminary injunction, Dr. Turk did not even reference the Little study. See Pl. Direct Evidence, Testimony of Dr. Amos Turk. Nevertheless, in his rebuttal affidavit dated September 28, 1994, Dr. Amos Turk appeared to depart from his prior deposition testimony and his lack of initial affidavit testimony when he stated that "the Little tests were not as probative as I had thought." See Pl. Rebuttal Evidence, Testimony of Dr. Amos Turk at ¶ 21. Moreover, during cross-examination, Dr. Turk vacillated on the probative value of the Little study, in clear contradiction of his prior sworn testimony. See Cross-Examination of Dr. Amos Turk, dated October 4, 1994. This Court is troubled by the internal inconsistencies with the respective written and oral testimony of Dr. Amos Turk regarding the Little study. Accordingly, this Court will entirely disregard the Little study and all of Dr. Amos Turk's testimony relating to it.

During trial, defendant S.C. Johnson's counsel requested that the Court participate in a sensory perception test. See Oral Argument of Lawrence I. Weinstein, dated October 4-5 17, 1994. Specifically, Mr. Weinstein desired this Court to perceive the alleged olfactory superiority of the Glade formulation as compared to pure baking soda when applied to cat urine. Id. The Court denied defense counsel's request on three bases. Id. First, a sensory perception experiment by the trier of fact cannot be replicated for appellate review unlike the viewing of trade dress configurations. Second, pursuant to FED R.EVID 403, "evidence may be excluded . . . by considerations of undue delay, waste of time, or needless presentation of cumulative evidence." Id. In the present action, both parties presented numerous expert witnesses, reports and studies regarding the olfactory efficacy of both the Glade formulations as well as baking soda. The Court's own subjective olfactory perception, which could be affected by a common cold, allergies, the air quality in the courtroom, or the particular pungency of the cat urine in question, is of little relevance in comparison. Moreover, participation in such a subjective test, which can neither be replicated nor cross-examined, could have the effect of making the Court a witness, contrary to FED.R.EVID. 605.

F. MARKET SURVEY EVIDENCE

1. The Bruno Ridgeway Survey

On June 29, 1994, Church Dwight's counsel contacted Joseph M. Ridgeway, President of Bruno Ridgeway Research Associates, Inc. (hereinafter "Bruno Ridgeway"), to conduct "an advertising comprehension study of a television commercial for Glade carpet deodorizer[s]." See Pl.Direct Evidence, Testimony of Joseph M. Ridgeway at ¶ 9. Mr. Ridgeway explained that

it quickly became apparent that the precise issue to be surveyed would require some sort of focused question which required the respondents to chose between two or more alternative meanings of the "five times better" claim. . . . [I]t became apparent that the simplest — and fairest — way of doing so was to give the respondents a choice between the interpretation urged by Church Dwight and the interpretation urged by S.C. Johnson. Of course, we also gave the respondent the option of choosing "neither" of those alternatives.

Id. at ¶¶ 12-13. The market survey was conducted at malls in Bridgeport, Connecticut; Greensboro, North Carolina; Oklahoma City, Oklahoma; Portland, Oregon; Philadelphia, Pennsylvania; and Peoria, Illinois from July 14, 1994 through July 20, 1994. See Pl.Direct Evidence, Testimony of Joseph M. Ridgeway at ¶¶ 16, 18. The mall interviewers surveyed 421 women 18 years or older, who had used either a carpet deodorizer or baking soda to deodorize their carpet within the last six months. Id. at ¶¶ 18-19. Questions 8, 9 and 10 were the only substantive questions of the Bruno Ridgeway survey. Id. at ¶¶ 14-15.

Question 8(a), known in market research parlance as an open-ended question, queried respondents: "What is the main idea that this commercial is communicating to you about Glade Carpet Potpourri?" See Pl.Trial Ex. 15 at 000829. Question 8(b) asked what "other ideas" were conveyed by the S.C. Johnson commercial. Id. Following question 8(b), the questionnaire instructed the interviewer as follows: "IF RESPONDENT SAYS GLADE IS 5X BETTER THAN BAKING SODA — CONTINUE WITH Q.9a. IF RESPONDENT SAYS GLADE IS BETTER BUT NOT BY HOW MUCH — ASK Q. 8c OTHERWISE — END INTERVIEW." Id. Question 8(c) queried "Did the commercial say how much better Glade Carpet Potpourri is than baking soda?" Id. If the respondent did not specifically mention "five times," the interviewer was instructed to end the interview. Id.

If in response to either question 8(b) or 8(c) the respondent specifically referenced the "five times better" claim, then the interviewer asked questions 9(a) and 9(b). Id. Question 9(a), also an open-ended question, stated: "In the commercial they said `Glade Carpet Potpourri is 5 times better than baking soda.' What does that mean to you?" Id. Question 9(b), again an open-ended question, asked: "What else does the idea that Glade is 5 times better than baking soda mean to you?" Id. Finally, the interviewer handed the respondent an index card with question 10 on it, which asked:

Which of the statements on this card best describes what this commercial communicates to you about Glade Carpet Potpourri?

A. 1()Glade Carpet Potpourri absorbs odor particles 5 times better than baking soda and reduces the odor you will smell better than baking soda, but not necessarily 5 times better.

B. 2()Glade Carpet Potpourri absorbs odor particles 5 times better than baking soda and reduces the odor you will smell 5 times better than baking soda.

C. 3()Neither of the above statements.

Id. Unlike questions 8(a), 8(b), 8(c), 9(a), and 9(b), question 10 was a closed-ended question in that it directed the respondent to choose one of three specific responses. See Pl.Rebuttal Evidence, Testimony of Susan S. McDonald at ¶¶ 16-18. Moreover, Joseph Ridgeway alternated responses A and B on the index card to insure statistical accuracy of this closed-ended question. See Cross-Examination of Joseph M. Ridgeway, dated October 17, 1994.

In response to question 8(b) and/or 8(c), 277 of the 421 respondents (66%) indicated that the commercial communicates that Glade is "five times better" than baking soda. Id. at ¶ 20. The survey instructed the interviewer that if the respondent mentioned "five times better" in question 8(b) or 8(c), then the interviewer was to ask the respondent questions 9(a) and 9(b). Id. at ¶ 21. Thus, the interviewers presented the 277 respondents with questions 9(a) and 9(b) which asked what the "five times better" claim meant to the respondent. Id. In response to question 9(a) and 9(b), "many respondents simply played back the five times better message but in slightly different language, while others said it meant that glade worked `five times' better (or at least `better') than baking soda." Id. The 277 respondents also were presented with the index card which contained closed-ended question 10. Id. at ¶ 22. In response to question 10, 72% of the respondents concluded that S.C. Johnson's commercial communicated that "Glade Carpet Potpourri absorbs odor particles 5 times better than baking soda and reduces the odor you will smell 5 times better than baking soda." Id. Joseph Ridgeway noted that "the survey was designed so that the key question was question 10, which asked the respondent to choose between the two alternative interpretations of the claim. On this score, the data overwhelmingly supports Church Dwight's interpretation of the claim." Id.

2. The Sorensen Surveys

In August 1994, counsel for defendant S.C. Johnson, retained Robert C. Sorensen of Sorensen Marketing/Management Corporation (hereinafter "Sorensen") to conduct a market research survey regarding the "five times better" claim. See Affidavit of Robert C. Sorensen at ¶¶ 10-12. Similar to the Bruno Ridgeway survey, the Sorensen survey interviewed "[w]omen 18 years of age or older who indicate[d] they had purchased and use a carpet deodorizer product two or more more times during the preceding nine months or believe they [would] purchase and use a carpet deodorizer product two or more times during the coming nine months." Id. at ¶ 15. The interviews were conducted at malls in New York, New York; Boston, Massachusetts: Sarasota, Florida; Kansas City, Missouri; and Houston, Texas. Id. at ¶ 16. Sorensen utilized two questionnaires in this survey: (1) the Screener Questionnaire, and (2) the Main Questionnaire. Id. at ¶ 23. "The Screener Questionnaire [merely] established that the survey respondent qualified" for the interview. Id.

The Main Questionnaire contained two open-ended questions which were asked following the respondent's viewing of S.C. Johnson's Glade commercial. See Def.Trial Exhibits, Volume II, Ex. 40 (Main Questionnaire). The first open-ended question asked "[n]ow please tell me the things this commercial is saying." Id. The second open-ended question asked "[y]ou just said . . . [at this point the interviewer was to repeat respondent's thoughts from the first question]. What do you understand that to mean?" Id. "This procedure was repeated for as many separate `thoughts' or numbered answers as were recorded in answer to Question la/b. . . ." See Affidavit of Robert C. Sorensen at ¶ 35. To tabulate the results of the 478 completed surveys, Sorensen employed a coding operation that assigned a specific number to eight survey perception themes. Id. at ¶ 27. Thereafter, a coding analyst checked each respondent's answers for the presence or absence of the designated themes and recorded the appropriate code number next to the answer on each questionnaire. Id. After tabulating the coded themes, Sorensen concluded that:

a net 58.2% of the total survey respondents in answer either to Questions 1a/1b or to Question 2, simply quote or cite one or more of the commercial's major "five times better" comparisons almost all of which offer "better", "tougher", "fresher", or "absorbs tough carpet odors better" advantages of Glade over baking soda, without respondent indicating she can/will be able to measure/tell this condition through her own sensory experience. . . . . 66.1% of the total survey respondents [employ] only a non-quantitative comparative advantage of Glade over baking soda.

Id. at ¶ 44.

In late September 1994, Sorensen conducted a second survey (hereinafter "second Sorensen survey") to determine the relative impact of two different television commercial statements on consumers' propensity to purchase Glade products. See Defendant's Trial Ex. 75; see also Second Affidavit of Robert C. Sorensen at ¶ 4. In this telephonic survey, conducted by PERT Survey Research of Bloomfield, Connecticut, a total of 373 "[w]omen 18 years of age and older who indicate[d] [that] they had purchased and used a carpet deodorizer product two or more times during the preceding nine months or believe they will purchase and use a carpet deodorizer product two or more times during the coming nine months." Id. at ¶ 10. In his second affidavit, Sorensen states that plaintiffs own sensory tests revealed that a malodorous "carpet treated with Glade on average smells 48% better than a carpet treated with baking soda at the 2 hour interval." See Second Affidavit of Robert C. Sorensen at ¶ 6. Thus, Sorensen wanted to survey consumers' purchasing reactions upon hearing one of the following two commercials: (1) "If your carpet or rug has an odor problem, it will smell five times better if you treat it with Glade Carpet Potpourri than it will if you treat it with baking soda," or (2) "If your carpet or rug has an odor problem, it will smell 48% better if you treat it with Glade Carpet Potpourri than it will if you treat it with baking soda." Id. at ¶¶ 8-9. In should be noted that approximately one-half of the respondents were asked statement one, while the other half were asked statement two. Id. at ¶ 11. Based upon the 373 telephonic interviews, Mr. Sorensen concluded that "[c]onsumers do not consider S.C. Johnson's `five times better' statement any greater incentive to buy Glade Potpourri than a `48% better' statement. . . ." Id. at ¶ 25.

III. CONCLUSIONS OF LAW AND FURTHER SPECIFIC FINDINGS OF FACT

PERMANENT INJUNCTION

The grant of a permanent injunction is an "extraordinary remedy, which should be granted only in limited circumstances." Frank's GMC Truck Center; Inc. v. General Motors Corp., 847 F.2d 100, 102 (3d Cir. 1988) (citing United States v. City of Philadelphia, 644 F.2d 187, 191 n. 1 (3d Cir. 1980)). The standard for a permanent injunction is essentially the same as for a preliminary injunction with the exception that the movant must demonstrate actual success on the merits rather than showing a likelihood of success on the merits. Amoco Production Co. v. Village of Gambell, 480 U.S. 531, 546 n. 12, 107 S.Ct. 1396, 1404 n. 12, 94 L.Ed.2d 542 (1987); see also Ciba-Geigy Corp. v. Bolar Pharmaceutical Co., Inc., 747 F.2d 844, 849 (3d Cir. 1984), cert. denied, 471 U.S. 1137, 105 S.Ct. 2678, 86 L.Ed.2d 696 (1985). Thus, a court must consider four factors when determining whether a permanent injunction should issue: (1) the moving party's success on the merits; (2) the probability of irreparable injury to the moving party in the absence of relief; (3) the potential harm to the non-moving party; and, if applicable, (4) the public interest. Fechter v. HMW Indus., Inc., 879 F.2d 1111, 1116 (3d Cir. 1989) (citing United States v. Price, 688 F.2d 204, 211 (3d Cir. 1982)); see also Opticians Ass'n of America v. Independent Opticians of America, 920 F.2d 187 (3d Cir. 1990).

A. SUCCESS ON THE MERITS

Section 43(a) of the Lanham Act, 15 U.S.C. § 1125(a), is broadly construed by courts to protect consumers from false advertising and from advertising which tends to falsely describe or represent products sold in commerce. L'Aiglon Apparel v. Lana Lobell, Inc., 214 F.2d 649 (3d Cir. 1954). Under section 43(a), there are two different theories of recovery for false advertising. Castrol Inc. v. Pennzoil, 987 F.2d 939, 943 (3d Cir. 1993). Specifically, a plaintiff challenging a competitor's advertisement must prove either that the advertisement is literally false on its face or "`the advertisement is literally true, but given the merchandising context, it nevertheless is likely to mislead and confuse consumers.'" Id. (quoting Johnson Johnson v. GAC Int'l, Inc., 862 F.2d 975, 977 (2d Cir. 1988)). Recently, in Johnson Johnson-Merck v. Rhone-Poulenc Rorer, 19 F.3d 125 (3d Cir. 1994), the Third Circuit noted:

If a plaintiff proves a challenged claim is literally false, a court may grant relief without considering whether the buying public was misled. A determination of literal falsity rests on an analysis of the message in context. 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 conveyed to the consumers. Public reaction is the measure of a commercial's impact.

Id. at 129. Thus, this Court must determine whether plaintiff Church Dwight has proven that defendant S.C. Johnson's "five times better" advertisements are literally false or are likely to mislead and confuse consumers.

1. Literally False Advertising

The test for literally false advertisements is quite simple: "[I]f a defendant's claim is untrue, it must be deemed literally false." Castrol Inc., 987 F.2d at 944. In the present action, defendant S.C. Johnson claims in various advertisements that its Glade carpet deodorizing products "absorb odors five times better than baking soda." Therefore, this Court must consider all the evidence presented to determine whether defendant S.C. Johnson's claim is literally false regarding its Glade regular formulation and its Wet'n Dry formulation.

A. Glade Regular Formulation

Defendant S.C. Johnson premises its "five times" superiority claim on a series of laboratory tests employing a Photoionization Analyzer. In fact, "Church Dwight agrees that this test is a good way to measure the absorption of volatile organic compounds, and does not contest the technical validity of the [S.C. Johnson] studies." See Pl.Trial Brief at 21. In a laboratory setting, using a minute amount of the regular Glade formulation, baking soda, and actual malodorants, defendant S.C. Johnson demonstrated that its regular formulation absorbs malodorant VOCs at least five times better than baking soda. See Affidavit of Dr. Peter E. Nott at ¶ 30. As defendant's expert, Dr. Peter E. Nott, stated in his affidavit "[t]he ratio of VOC reduction in Glade's favor was not merely 5 to 1 but infinite, as the baking soda actually increased the VOCs of cat urine, while Glade reduced the VOCs." Id.

While defendant S.C. Johnson successfully demonstrated a five-fold superiority in laboratory testing, its various advertisements merely claim that it "absorbs odors five times better than baking soda" and do not inform consumers that this claimed superiority is based solely on laboratory testing. When viewed as a whole and in context, defendant S.C. Johnson's "five times better" claims in the television commercial, the magazine advertisement and the product labeling each must be fairly understood as saying that the Glade products work "five times better" at absorbing odors in the real world. In fact, in the television commercial, defendant S.C. Johnson proclaims that its new Glade products are "five times fresher" than baking soda. See Pl. Trial Ex. 8. Moreover, in the magazine advertisement, defendant S.C. Johnson states that the new Glade products have five time the absorbing power "to keep your home smelling fresh and clean." Id. at Ex. 9. The language in S.C. Johnson's advertisements surely intimates that its "five times better" claim can actually be replicated in the real word, leaving a consumer's home "five times fresher." Id. at Ex. 8. As the Third Circuit held in Castrol Inc., "the test for literal falsity is simpl[e]; if a defendant's claim is untrue, it must be deemed literally false. . . . In this case, [defendant] made a claim of superiority, and when tested, it proved false. Hence, under this standard. the district court correctly found literal falsity." See Castrol Inc., 987 F.2d at 944. Since defendant S.C. Johnson fails to notify consumers that its superiority claim is derived solely from laboratory tests, this Court must review the evidence presented to determine whether this claim can be replicated in normal consumer usage. In other words, that defendant's claimed superiority is true in a laboratory setting, does not rescue it from literal falsity should this Court conclude that the laboratory claim has no practical equivalent in the real world.

During cross-examination, plaintiff's counsel, Thomas C. Morrison, questioned Dr. Peter E. Nott on the minute amount of Glade formulation and malodorants used in S.C. Johnson's laboratory studies. See Cross-Examination of Dr. Peter E. Nott, dated October 6, 1994. Dr. Nott acknowledged that for the photoionization studies he used 10 grams of Glade formulation and 15 microliters of cat urine. Id. When asked to quantify that amount of cat urine, Dr. Nott testified that 15 microliters, which is .015 milliliters, is approximately equivalent to .375 of a medicine drop. See Plaintiffs Evidence, Ex. 45. Dr. Nott acknowledged that the typical volume of cat urine per incident is approximately 25 milliliters. See Cross-Examination of Dr. Peter E. Nott, dated October 6, 1994. Thus, the amount of cat urine employed in defendant S.C. Johnson's study was roughly equivalent to 1/1667 of the volume of a typical cat incident. Id. Continuing in this vein, Mr. Morrison questioned Dr. Nott regarding the amount of actual Glade product that would be required to replicate the S.C. Johnson study for the average cat incident of 25 milliliters. Id. It was ultimately determined that for a cat incident of 25 milliliters, 16,670 grams of Glade product — approximately 37 one pound cans of product — would be required to replicate the study in the correct proportion. Id. Indeed, one wonders if absorption, adsorption or neutralization have any relevance to such a scenario; 37 pounds of product would form a substantial pyramid over 25 milliliters of cat urine. Accordingly, this Court finds that all of defendant S.C. Johnson's advertising, which claims that Glade regular formulation "absorbs odors five times better than baking soda," is literally false.

Moreover, in the context of pharmaceutical false advertising claims, courts have recognized that in vitro or laboratory testing is not necessarily representative of in vivo performance unless a specific correlation has been established. See United States v. Premo Pharmaceutical Laboratories, 511 F. Supp. 958, 985 (D.N.J. 1981); see also Johnson Johnson-Merck v. Rhone-Poulenc Rorer, 19 F.3d at 127 (Third Circuit discussed in vitro and in vivo testing and noted that information derived from in vivo studies are "more useful to consumers than the results of laboratory studies. . . ."). In vitro is defined as "in a test tube or other artificial environment," while in vivo is defined as "in the living body of a plant or animal." Webster Third New International dictionary 1190 (1976). As the district court recognized in Premo Pharmaceutical Laboratories, "[ i] n vitro tests do not provide a basis for conclusions . . . unless a correlation has been made between the results of those studies and the results of an in vivo study for the drug." Premo Pharmaceutical Laboratories, 511 F. Supp. at 985. As such, pharmaceutical manufacturers generally cannot assert a claim based on in vitro studies unless it has established a specific correlation between those studies and "the results of an in vivo study for the [specific] drug." Id. ⁴

4 In Johnson Johnson-Merck, the Third Circuit recognized that in the limited context of over-the-counter antacids, "the test on which advertisers rely for claims of [neutralization] strength does not measure the operation of the antacids in the human body ( in vivo testing); rather, it measures their operation in glass beakers of acid in the laboratory ( in vitro testing). . . . [because] the FDA concluded [that] human testing would be too laborious." Johnson Johnson-Merck, 19 F.3d at 127.

By analogy, this Court will extend the in vitro/in vivo pharmaceutical correlation requirement to the present action. In the context of this litigation, the term in vitro connotes research conducted in a laboratory setting, while the term in vivo connotes a practical testing application designed to closely replicate the actual, intended use by the consumer. While defendant S.C. Johnson demonstrated a five-fold superiority in in vitro testing, its "five times better" advertisements make absolutely no distinction between in vitro and in vivo testing. When viewing the advertisements, a consumer can reasonably perceive that defendant S.C. Johnson's "five times better" assertion is actually an in vivo claim — namely, something that can be replicated in the real world. This Court finds, however, that defendant S.C. Johnson has failed to provide any correlation between its in vitro studies and its in vivo "five times better" claim. Accordingly, for this reason as well, the Court finds that defendant S.C. Johnson's "five times better" advertising claim is literally false.

B. Glade Wet'n Dry Formulation

As stated previously, defendant S.C. Johnson's Wet'n Dry formulation is composed of 94% baking soda, 4% sodium sulfate. 1% aluminum silicate (zeolite), and 1% fragrance. See Pl. Trial Ex. 21 at 1-2. Although the formulation of the Glade Wet'n Dry is quite different from the Glade regular formulation, defendant S.C. Johnson nevertheless advertises its Wet'n Dry formulation as "five times better than baking soda." See Pl. Trial Exs. 5, 6, 8, and 9. No where on the product packaging or in any of the advertisements does defendant S.C. Johnson inform consumers that its Wet'n Dry formulation is composed almost entirely of baking soda. Id. Instead, on the Wet'n Dry package, defendant advertises that the product "ABSORBS ODORS 5 TIMES BETTER than BAKING SODA." Id.

In affidavit testimony, Dr. Peter E. Nott, S.C. Johnson's research manager, informed the Court:

The results of the Electronic Nose test showed that [the Wet'n Dry] formulation was also effective in reducing cat urine VOCs, which would indicate that the cat urine VOC reduction ratio in favor of that formulation was also infinite because, as discussed above, baking soda consistently failed to decrease cat urine VOCs.

See Affidavit of Peter E. Nott at ¶ 36. Defendant S.C. Johnson never performed head-to-head studies comparing the efficacy of its Wet'n Dry formulation and pure baking soda. Id. Thus, defendant S.C. Johnson cannot verify that through laboratory testing, its Wet'n Dry formulation worked "five times better than baking soda." Id. Additionally, this Court is unpersuaded by Dr. Nott's assumption that the "cat urine VOC reduction ratio in favor of [the Wet'n Dry] formulation was also infinite because, as discussed above, baking soda consistently failed to decrease cat urine VOCs." Id. Dr. Nott's assumption, however, fails to consider the fact that the Wet'n Dry formulation is composed of 94% baking soda; the same chemical agent Dr. Nott testified actually increases cat urine VOCs. Id. at ¶ 31. Accordingly, this Court finds that defendant S.C. Johnson's "five times better claim," in reference to its Wet'n Dry formulation, also is literally false.

2. CONSUMER SURVEYS IN FALSE ADVERTISING LITIGATION

Alternatively, even if defendant S.C. Johnson's "five times better" claim was not literally false, it would have to be enjoined as deceptive or misleading. Whether an advertising claim "is deceptive or misleading, . . . depends on the message that is conveyed to consumers." Johnson Johnson-Merck, 19 F.3d at 129. The success of a plaintiff's false advertising claim will turn "on the persuasiveness of a consumer survey." Id. at 129-30. A Lanham Act plaintiff "cannot obtain relief by arguing how consumers could react; it must show how consumers actually do react." Sandoz Pharmaceuticals Corp. v. Richardson-Vicks, Inc., 902 F.2d 222, 228 (3d Cir. 1990). Therefore, this Court must evaluate the two consumer surveys presented in this litigation — the Bruno Ridgeway survey and the Sorensen survey — to "determine whether the public was, in fact, misled." Id. at 130.

Although both surveys purport to evaluate consumer reactions to defendant S.C. Johnson's "five times better" claim, they do so in extremely different fashions. In Pittsburgh Press Club v. United States, 579 F.2d 751, 758 (3d Cir. 1978), the Third Circuit stated that the proponent of a consumer survey has the burden of establishing that it was conducted in accordance with accepted principles of survey research. Id. The Third Circuit noted that generally accepted survey principles include the following:

A proper universe must be examined and a Representative sample must be chosen; the persons conducting the survey must be experts; the data must be properly gathered and accurately reported. It is essential that the sample design, the questionnaires and the manner of interviewing meet the standards of objective surveying and statistical techniques. Just as important, the survey must be conducted independently of the attorneys involved in the litigation. The interviewers or sample designers should, of course, be trained, and ideally should be unaware of the purposes of the survey or the litigation. A fortiori, the Respondents should be similarly unaware.

Id. (citation omitted). Thus, this Court must first ascertain whether the Bruno Ridgeway survey and the Sorensen survey were conducted in accordance with accepted principles of survey research. It should be noted, however, that "it is relatively easy for one ad expert to criticize a survey done by another expert." U-Haul International, Inc. v. Jartran, Inc., 522 F. Supp. 1238, 1252 (D.Ariz. 1981), aff'd, 681 F.2d 1159 (9th Cir. 1982). As J. Thomas McCarthy stated in his off-cited treatise, McCarthy on Trademarks and Unfair competition:

One must keep in mind that there is no such thing as a "perfect" survey. The nature of the beast is that it is a sample, albeit a scientifically constructed one. One must remain cognizant of the fact that, as Judge Anderson said of the survey evidence in the famous "Thermos" case, "any conclusion in this area cannot be reduced to a figure of unimpeachable accuracy but must, at best, be an approximation." Like any scientific method related to statistics in the social sciences, every survey, no matter how carefully constructed and conducted, has some potential flaws somewhere.

The proper approach is to view such evidence with some understanding of the difficulty of devising and running a survey and to use any technical defects only to lessen evidentiary weight, not to reject the results out-of-hand.

4 McCARTHY ON TRADEMARKS AND UNFAIR COMPETITION § 32:50[5], at 32-228-32-229 (3d ed. 1994). Finally, it should be noted that "[s]urveys are statistical evidence and are an aid to, not a substitute for, independent evaluation by a judge or jury." Id. § 32:54[1][b][ii] at 32-238.

The two surveys in question employed substantially similar survey universes. The Bruno Ridgeway survey interviewed a total of 421 respondents at six malls throughout United States. See Pl. Direct Evidence, Testimony of Joseph M. Ridgeway at ¶¶ 16, 18-19. The 421 Bruno Ridgeway respondents were women 18 years of age or older who had either used a carpet deodorizer or baking soda to deodorize their carpet within the last six months. Id. at ¶¶ 18-19. The Sorensen survey interviewed a total of 478 respondents at five malls throughout the United States. See Affidavit of Robert C. Sorensen at ¶¶ 16, 27. The 478 Sorensen respondents were "[w]omen 18 years of age or older who indicate[d] they had purchased and used a carpet deodorizer product two or more times during the preceding nine months or believe they [would] purchase and use a carpet deodorizer two or more times during the coming nine months." Id. at ¶ 15.

Dr. Ivan Ross, defendant's survey expert hired to critique the Bruno Ridgeway survey, stated in his affidavit:

It is always incorrect to define the universe for a false advertising study as consisting only of users, as distinct from purchasers or purchase decision-makers. If an advertisement is deceptive, it presumably would have impact on those individuals who make the actual purchase or purchase decision. Those individuals may also be product users, but not necessarily. Additionally, the universe should be defined as including prospective purchasers or purchase decision-makers, not just past purchasers. The Bruno and Ridgeway survey failed on both of those counts. . . . In addition to this fundamental flaw, no evidence is provided that the respondents, all of whom were women, are reasonably representative of the age of women known to be product category purchasers and users. In sum, the [Bruno Ridgeway] survey does not meet any conceivable criteria to be a representative sample drawn from a meaningfully defined universe.

See Affidavit of Dr. Ivan Ross at ¶ 10. The only substantive difference this Court can detect between the Bruno Ridgeway survey universe and the Sorensen survey universe is the exclusion of prospective purchasers in the former and the presence of prospective purchasers in the latter. Dr. Susan S. McDonald, plaintiff's survey expert hired to critique the Bruno Ridgeway and Sorensen surveys, stated that Dr. Ross's criticisms of lack of prospective purchasers in the Bruno Ridgeway survey "is nonsense. In marketing, there is no better predictor of future behavior than past behavior, especially where, as here, the product category is characterized by frequent purchases." See Pl. Rebuttal Evidence, Testimony of Susan S. McDonald at ¶ 46. This Court is persuaded by Dr. McDonald's affidavit testimony and concludes that, in a consumer survey dealing with a product that is sold to approximately 18 million households annually, see Affidavit of Catherine Clemency at ¶ 4, past purchasers' behavior provides the most accurate forecast of future behavior. Accordingly, this Court finds that both surveys in question examined a proper universe when conducting the interviews.

This Court will now address the substantive questions of both surveys. Purporting to rely on Johnson Johnson-Merck v. Rhone-Poulenc Rorer, 19 F.3d 125 (3d Cir. 1994), defendant S.C. Johnson alleges that questions 8(c), 9(a) and 9(b) of the Bruno Ridgeway survey are "very leading" and therefore should be dismissed. See Affidavit of Dr. Ivan Ross at ¶¶ 15-18; see also Def. Post-Trial Br. at 9. In Johnson Johnson-Merck, the Third Circuit held that the district court was not clearly erroneous when it determined that a series of three Bruno Ridgeway open-ended, consumer survey questions in that case were "very leading" — i.e., they influenced respondents to give a particular answer — and therefore should be disregarded. Id. at 136. Although the three suspect questions in the Johnson Johnson-Merck case are somewhat similar to survey questions 9(a) and 9(b) in this case, id. at 133 n. 12, 135 n. 15, this Court cannot conclude that the Third Circuit held as a matter of law that all such open-ended, consumer survey questions are leading and therefore should be disregarded. Instead, the opinion stands for the proposition that, in false advertising action, the evaluation of consumer surveys is "within the sound discretion of the district court." Id. at 135. Thus, the Third Circuit held that the district court in Johnson Johnson-Merck was not clearly erroneous when it dismissed specific survey questions as "very leading." Id. at 135-36. Because our Court of Appeals has explicitly held that the evaluation of a consumer survey lies within the sound discretion of the district court, we must evaluate questions 8(c), 9(a) and 9(b) of the Bruno Ridgeway survey in light of the testimony and evidence presented in this case. Id. at 135.

To evaluate whether questions 8(c), 9(a) and 9(b) of the Bruno Ridgeway survey are impermissibly leading, we will consider the affidavit and cross-examination testimony of Dr. Susan S. McDonald, plaintiff's survey expert, and Dr. Ivan Ross, defendant's survey expert. Dr. Ivan Ross characterizes these questions as a "sledge hammer" or a "bludgeoning" approach. See Affidavit of Dr. Ivan Ross at ¶ 19. It is Dr. Ross's testimony that because question 8(c) queries respondent how many times better Glade Carpet Potpourri is than baking soda, followed by questions 9(a) and 9(b) which both reference the "five time better" claim, a respondent is "`shove[d]' toward the answer or thought that the interviewer is seeking." Id.

This Court is unpersuaded by Dr. Ross's criticisms of questions 8(c), 9(a) and 9(b). The Bruno Ridgeway survey was crafted to specifically test how consumers reacted to and understood defendant S.C. Johnson's "five times better" claim. As mentioned previously, in the 30-second audio segment of the commercial, defendant S.C. Johnson makes the following 5 claims: (1) "Glade . . . is now five times better than baking soda"; (2) "It's five times tougher"; (3) "absorbs tough carpet odors, five times better than baking soda"; (4) "It's five times fresher"; and (5) "It's five times better than baking soda." See Pl.Trial Ex. 8. As also noted supra, defendant S.C. Johnson reinforces its five-fold claim through the use of the following visual elements: (1) five men, sitting around a dining room table, smoking cigarettes in the third frame, followed by five empty chairs around the same dining room table in the fourth frame, and (2) five golden retriever puppies allegedly urinating on a living room carpet in the seventh frame, followed by three puppies in the eighth frame, only one puppy in the ninth frame, and no puppies remaining in the tenth frame in which a word overlay appears stating "Five Times Better Than Baking Soda." Id. As a result of defendant S.C. Johnson's bombardment of its "five times better" claim in the commercial, this Court finds that questions 8(c), 9(a) and 9(b) are relevant inquires. The Court will consider each of these questions in turn.

Question 8(c) merely asked a respondent if the commercials quantified the claimed superiority of Glade Carpet Potpourri. See Pl. Trial Ex. 15 at 000829. In Johnson Johnson-Merck v. Rhone-Poulenc Rorer, No. 91-7099, 1993 WL 21239 (E.D.Pa. Jan. 29, 1993), aff'd, 19 F.3d 125 (3d Cir. 1994), the district court criticized the open-ended questions contained in that survey because the questions failed to incorporate a "filter mechanism." Id. at *19. "A survey should be properly `filtered' to screen out those who got no message from the commercial." Id. (citing American Home Products v. Johnson Johnson, 654 F. Supp. 568, 590 (S.D.N.Y. 1987), aff'd, 848 F.2d 34 (2d Cir. 1988)). In the present case, however, questions 8(b) and 8(c) employ 3 filter mechanisms to "filter" out respondents who got no message from the commercial. See Pl. Trial Ex. 15 at 000829.

Following question 8(b), the interviewer is instructed to end the interview if the respondent makes no mention of Glade superiority over baking soda. Id. Furthermore, question 8(c) is asked only if a respondent indicates, in response to question 8(b), that the commercial communicates that "Glade is better but not by how much." Id. If when asked question 8(c) — "Did the commercial say how much better Glade Carpet Potpourri is than baking soda?" — the respondent answers either "no" or "don't know," the interviewer is instructed to end the interview. Id. In fact, if in response to question 8(c) the respondent does not specifically mention the "five times better" claim, the interviewer is again instructed to end the interview. Id. This Court finds that these 3 filter mechanisms are effective means of screening out respondents who got no message from defendant's commercial.

Additionally, this Court finds that questions 9(a) and 9(b) are relevant queries based on the repetitive nature of defendant S.C. Johnson's "five times better" claim communicated in the commercial. See Pl. Trial Ex. 8. A respondent would not even be asked questions 9(a) and 9(b) unless she specifically recalled the "five times better" claim in response to either 8(b) or 8(c). See Pl. Trial Ex. 15 at 000829. If a respondent is able to recall the "five times better" claim, then and only then is respondent asked questions 9(a) and 9(b) which state: "In the commercial, they said `Glade Carpet Potpourri is 5 times better than baking soda.' What does that mean to you?" and "What else does the idea that Glade is 5 times better than baking soda mean to you?" Id. Accordingly, this Court finds that because of the 3 filter mechanisms employed in questions 8(b) and 8(c), questions 9(a) and 9(b) are relevant inquires and are not leading questions when attempting to evaluate consumers' perception of defendant's commercial.

Another substantive difference between the Bruno Ridgeway and Sorensen surveys is the use of closed-ended survey questions as opposed to open-ended survey questions. Although the Bruno Ridgeway survey employs both open-ended and closed-ended questions, Dr. Ross's predominant criticism of the Bruno Ridgeway survey involves question 10, which is a closed-ended question. As explained previously, a closed-ended question directs a respondent to choose one of a series of specific responses in a multiple-choice format. See Pl. Rebuttal Evidence, Testimony of Susan S. McDonald at ¶¶ 16-18. As noted previously, question 10 asked respondents:

Which of the statements on this card best describes what this commercial communicates to you about Glade Carpet Potpourri?

A. 1()Glade Carpet Potpourri absorbs odor particles 5 times better than baking soda and reduces the odor you will smell better than baking soda, but not necessarily 5 times better.

B. 2()Glade Carpet Potpourri absorbs odor particles 5 times better than baking soda and reduces the odor you will smell 5 times better than baking soda.

C. 3()Neither of the above statements.

See Pl. Trial Ex. 15 at 000829. Defendant S.C. Johnson's specific criticisms of question 10 are two-fold: (1) the Bruno Ridgeway survey failed "to use a control question to subtract out the `noise' inherently created by a multiple choice format. . . .", and (2) the Bruno Ridgeway survey did not include among the possible answers to question 10 either the commercial's literal meaning or the interpretation advanced by S.C. Johnson. See Def. Post-Trial Br. at 14-20. Moreover, in their post-trial brief and throughout the trial, defense counsel continually referred the Court to the district court decision in Johnson Johnson-Merck v. Rhone Poulenc Rorer, No. 91-7009, 1993 WL 21239 (E.D.Pa. Jan. 29, 1993), aff'd, 19 F.3d 125 (3d Cir. 1994), and the district court decision in Johnson Johnson-Merck v. Smithkline Beecham Corp., No. 91-0960, 1991 WL 206312 (S.D.N.Y. Oct. 1, 1991), aff'd, 960 F.2d 294 (2d Cir. 1992), for the general proposition that open-ended consumer survey questions are far more probative and reliable indicators of consumer comprehension of advertising messages than are closed-ended questions. See Def. Post-Trial Br. at 8. The Court will first address defendant's general criticism of closed-ended questions, before addressing defendant's specific criticisms of question 10.

This Court was particularly persuaded by the affidavit testimony and cross-examination testimony of plaintiff's survey expert, Dr. Susan S. McDonald, who concentrates her practice in non-litigation related consumer research. See Cross-Examination of Dr. Susan S. McDonald, dated October 17-18, 1994. In her affidavit and at trial, Dr. McDonald testified that in non-litigation consumer research, "[t]he overwhelming majority of consumer research surveys rely entirely or primarily on closed-ended questions when quantitative . . . analysis is the objective." See Pl. Rebuttal Evidence, Testimony of Dr. Susan S. McDonald at ¶ 18. During trial, Dr. McDonald noted that companies generally employ closed-ended questions when conducting marketing research concerning particular products. See Cross-Examination of Dr. Susan S. McDonald dated October 17 and 18, 1994. Dr. McDonald explained that "closed-ended questions do a much better job of eliciting information that can be systematically tallied." See Pl. Rebuttal Evidence, Testimony of Dr. Susan S. McDonald at ¶ 18. Dr. McDonald explained that companies therefore use closed-ended survey questions to make business decisions that directly impact upon a company's financial situation. See Cross-Examination of Dr. Susan S. McDonald, dated October 17-18, 1994.

Dr. McDonald noted that "in the context of litigation, some courts have written decisions casting doubt on the appropriateness of directed questions," while in a non-litigation context, companies "make decisions involving multi-million dollars based on research involving closed-ended, or directed, questions." See Pl. Rebuttal Evidence, Testimony of Dr. Susan S. McDonald at ¶ 26; see also Cross-Examination of Dr. Susan S. McDonald, dated October 17-18, 1994. Ultimately, Dr. McDonald opined that the type of consumer survey questions companies employ when making multi-million dollar consumer marketing decisions are the most accurate and reliable measure of actual consumer perception. Id.

This Court agrees. If "responses to unaided, open-ended questions are [such] highly reliable and unbiased evidence of consumer comprehension of adverting messages," see Def. Post-Trial Br. at 8, then why do a vast majority of companies favor closed-ended questions when making multi-million dollar consumer marketing decisions? In a Lanham Act false advertising case, it is crucial that the survey evidence presented to the trier of fact most closely replicates consumers' "real world" perceptions of the contested advertisement. Accordingly, because marketers regularly employ closed-ended consumer survey questions when making multimillion dollar consumer marketing decisions, this Court finds it reasonable in the context of the present case to rely on unbiased, closed-ended questions when adjudicating a false advertising claim.

Unlike the Bruno Ridgeway survey, the Sorensen survey did not include any closedended questions. Instead, the Sorensen survey had only two substantive, open-ended questions which stated: "Now, please tell me the things this commercial is saying. Anything else?" See Def. Trial Exhibits, Volume II, Ex. 40 (Main Questionnaire). After reviewing the Sorensen survey questions, Dr. McDonald stated that "it is merely a recall test of the commercial which, like any recall test, yields unfocused results." See Pl. Rebuttal Evidence, Testimony of Dr. Susan S. McDonald at ¶ 28. Moreover, Dr. McDonald noted that there were numerous ambiguities and discontinuities in the open-ended responses which resulted in coding problems. Id. at ¶ 35. In conclusion, Dr. McDonald stated that the Sorensen coding scheme was "Byzantine" and did not accurately measure consumers' perceptions of defendant's commercial. Id. at ¶¶ 36-37. The Third Circuit has recognized that "`[t]he probative value of a consumer survey is a highly fact-specific determination and a court may place such weight on survey evidence as it deems appropriate.'" Johnson Johnson-Merck, 19 F.3d at 134 (citations omitted). Because in the context of the present case, this Court finds it appropriate to place great weight on unbiased, closed-ended survey questions when measuring consumers' perceptions, this Court will accord the Sorensen survey little weight.

We will now consider defendant's expert's particular criticisms of question 10 of the Bruno Ridgeway survey. It is worth noting that, unlike Dr. Susan S. McDonald, defendant's expert, Dr. Ivan Ross, concentrates almost exclusively in consumer research for litigation purposes. See Cross-Examination of Dr. Ivan Ross, dated October 18, 1994. In fact, Dr. Ross testified that he spends relatively little time conducting consumer research surveys for non-litigation purposes. Id. Dr. Ross criticized question 10 because: (1) it failed to use a control question to "filter" out respondents who got no message from the commercial; and (2) it allegedly failed to include, among its possible answers, either the commercial's literal meaning or the interpretation advanced by S.C. Johnson. See Def. Post-Trial Br. at 14-20. Dr. Ross's first criticism is easily disposed of because it is clear that Dr. Ross reviewed question 10 in a vacuum and not in the context of the entire Bruno Ridgeway survey. Specifically, this Court has already found that the Bruno Ridgeway survey employed not one, but three filter mechanisms which effectively screened out respondents who perceived no message from defendant's commercial.

Before addressing Dr. Ross's second criticism, this Court must determine the parties respective positions in regards to defendant's commercial. Based on the repeated "five times better" and "five times fresher" claim in defendant's commercial, plaintiff Church Dwight claims that consumers will equate "absorbs odors five times better" with a fivefold sensory experience. In contrast, defendant S.C. Johnson claims that although its Glade products absorb VOCs at least five times better than baking soda, consumers will not equate this claimed superiority with a five-fold sensory experience. Thus, the Bruno Ridgeway survey was designed to test consumers' perceptions relative to the parties' respective interpretations of the "five times better" claim. In the actual commercial, defendant uses the phrase: "It's five times fresher." See Pl. Trial Ex. 8. This Court finds that defendant's use of the particular adjective "fresher" connotes the olfactory efficacy of the Glade products as compared to baking soda. As such, this Court concludes that the choices presented in question 10 of the Bruno Ridgeway survey accurately represent the parties' respective positions concerning defendant's "five times better" claim.

In response to question 10, 72% of the respondents chose plaintiff Church Dwight's interpretation of the commercial. See Pl. Direct Evidence, Testimony of Joseph M. Ridgeway at ¶ 22. Specifically, respondents understood defendant's commercial as communicating: "Glade Carpet Potpourri absorbs odor particles 5 times better than baking soda and reduces the odor you will smell 5 times better than baking soda." Id. As the various sensory tests would indicate and as defendant S.C. Johnson repeatedly concedes, the reformulated Glade products were not capable of reducing the perceived odor a consumer would smell by five times. See, e.g., Reply Affidavit of Catherine Clemency at ¶ 7. Thus, the fact that 72% of respondents understood defendant's commercial to mean that the various Glade products were capable of reducing the odor a consumer smelled by five times, is clear evidence that defendant S.C. Johnson's commercial is deceptive and misleading. In fact, "[c]ourts have found survey figures to be sufficient when they revealed that 21 percent to 34 percent, `over 25 percent,' and 33 percent of respondents received a misleading message from the ad." 4 McCARTHY ON TRADEMARKS AND UNFAIR COMPETITION § 32:54[4], at 32-249 (3d ed. 1994) (citations omitted). Accordingly, this Court concludes that defendant S.C. Johnson's commercial violates section 43(a) of the Lanham Act, 15 U.S.C. § 1125(a).⁵

5 In its post-trial brief, defendant S.C. Johnson claims that "a plaintiff's claim that more than one nonidentical advertisement is deceptive requires a survey that tests consumer perception of each advertisement." See De. Post-Trial Br. at 4-5. Although plaintiff Church Dwight failed to conduct consumer surveys regarding the packaging and print advertisements, this Court need not address this issue because it has already found that those advertisements are "literally false."

B. IRREPARABLE INJURY

Although plaintiff Church Dwight has proven success on the merits, it must also establish that it has suffered irreparable injury. Under section 43(a) of the Lanham Act, "[i]rreparable injury does not require diversion of actual sales and it can include the loss of control of reputation, loss of trade, and loss of goodwill." W.L. Gore Assoc., Inc. v. Totes Inc., 788 F. Supp. 800, 810 (D.Del. 1992) (citing Opticians Ass'n of America v. Independent Opticians of America, 920 F.2d 187, 195 (3d Cir. 1990) (quoting 2 J. McCARTHY, TRADEMARKS AND UNFAIR COMPETITION § 30:18 (2d ed. 1984))). Clearly, defendant S.C. Johnson's false advertisements, which directly attack baking soda's efficacy as a deodorant, have caused plaintiff Church Dwight to lose control of its reputation and goodwill. Id. Accordingly, plaintiff has successfully demonstrated that it will suffer irreparable harm unless defendant is permanently enjoined from its false advertising.

C. BALANCING OF THE EQUITIES

"In deciding whether injunctive relief is appropriate, the third task a trial court must undertake is to balance the hardships to the respective parties." Opticians Ass'n, 920 F.2d at 197. Because this Court has already determined that plaintiff Church Dwight will suffer irreparable injury unless an injunction issues, this Court need only consider whether the issuance of a permanent injunction will unduly harm defendant. Id. Catherine Clemency, defendant S.C. Johnson's Category Manager for its Air Care products, testified in her affidavit that "[a]lthough restaged Glade has been in broad-scale distribution for approximately the past five months, we have not yet seen an increase in market share. Indeed, the latest market share data shows Church and Dwight's market share rising, while S.C. Johnson's market share is down." See Affidavit of Catherine Clemency at ¶ 22. This Court concludes that based on the fact that S.C. Johnson's false advertisements have yet to be effective, a permanent injunction will not unduly burden defendant.

D. PUBLIC INTEREST

Section 43(a) of the Lanham Act was promulgated to protect consumers from false and deceptive advertising. 15 U.S.C. § 1125(a). "The public has a right not to be deceived or confused." W.L. Gore Assoc., Inc., 788 F. Supp. at 813 (citations omitted). Because defendant S.C. Johnson's false advertisements actually deceive and confuse the consuming public, there is sufficient public interest to permanently enjoin defendant from continuing in its deceptive advertising practices. Accordingly, because plaintiff Church Dwight has satisfied the four requirements necessary for the issuance of a permanent injunction, and this Court will permanently enjoin defendant S.C. Johnson from advertising its products as "five times better than baking soda." Plaintiff shall submit an appropriate form of order within 10 days.