UNITED STATES DISTRICT COURT NORTHERN DISTRICT OF CALIFORNIA
RANDALL HEAGNEY, et al., Case No. 23-cv-00687-VC
Plaintiffs, ORDER ON MOTION FOR CLASS v. CERTIFICATION, MOTION TO EXCLUDE, AND MOTION FOR JOHN PAUL MITCHELL SYSTEMS, LEAVE TO AMEND Defendant. Re: Dkt. Nos. 120, 121, 124
The motion for class certification is granted, although with respect to a narrower class than proposed by the plaintiffs. JPMS’s motion to exclude the plaintiffs’ expert witness and the plaintiffs’ motion for leave to amend are denied. This order assumes the reader’s familiarity with the facts, governing legal standards, and arguments made by the parties. Class certification. A class of all California residents who between May 1, 2015, and January 1, 2020, purchased, directly from JPMS or through an authorized third-party retailer or salon, a JPMS hair care product branded as Paul Mitchell, Clean Beauty, Tea Tree, MITCH, Awapuhi Wild Ginger, Neuro, or MVRCK, is certified. Plaintiff Randall Heagney is appointed class representative. Material differences in state laws preclude certification of a nationwide class. With respect to the plaintiffs’ UCC claims, “there are material conflicts between California warranty law and the warranty law of the other forty-nine states.” In re Hitachi TV Optical Block Cases, 2011 WL 9403, at *6 (S.D. Cal. Jan. 3, 2011). There are also differences between California’s consumer protection statutes and those of other states, including the states the plaintiffs (other than Heagney) live in: Virginia, for instance, does not allow class actions, while Texas requires pre-suit notice to the defendant. See Gustafson v. BAC Home Loans Servicing, LP, 294 F.R.D. 529, 539 (C.D. Cal. 2013); Darisse v. Nest Labs, Inc., 2016 WL 4385849, at *10 (N.D. Cal. Aug. 15, 2016). Because the relevant state laws differ materially, classmembers’ claims must be governed by the laws of their home states. See Mazza v. American Honda Motor Co., 666 F.3d 581, 593–94 (9th Cir. 2012). Abstract “reliance” or “non-reliance” subclasses cannot be certified because “[c]hoice of law requires determining the single state’s substantive law that applies to a particular plaintiff’s claim.” Cimoli v. Alacer Corp., 587 F. Supp. 3d 978, 988 (N.D. Cal. 2022) (emphasis added). Nor can subclasses be certified for the states where the other plaintiffs live because the plaintiffs did not bring claims under the laws of those states—and as discussed below, their motion for leave to amend is denied. Because Heagney is the only named plaintiff to whom California law applies, the class claims can only proceed as to the representation he saw and says he relied on—that is, the “no animal testing” statement, as well as the substantially similar “JPMS does not conduct or endorse animal testing” statement. Cf. Bazarganfard v. Club 360 LLC, 344 F.R.D. 411, 426 & n.14 (C.D. Cal. 2023). There is no evidence, and the plaintiffs do not allege, that he saw the “pioneer in cruelty-free haircare” representation. Moreover, he stopped using JPMS products “around 2020,” and that slogan wasn’t used until that year. While it’s a close call based on the evidence the plaintiffs present about industry usage and consumers’ understanding of the terms, that statement is too different from “no animal testing” to be considered substantially the same representation. Cf. Conde v. Sensa, 2019 WL 4277414, at *8 (S.D. Cal. Sep. 10, 2019). For that reason, and because the plaintiffs have not shown how it could be determined which representation was seen by class members who purchased the products during the period that both statements were in the market, it is appropriate for the class period to end in January 2020, when the “pioneer” statement began to be used. The rest of JPMS’s challenges to certification fail. First, although there isn’t adequate proof of classwide exposure to the marketing statements, there is evidence that the targeted representations were visible enough and “consistently made on a product’s label.” Krommenhock v. Post Foods, LLC, 334 F.R.D. 552, 565 (N.D. Cal. 2020). Second, reliance can be inferred on a classwide basis because there is sufficient evidence of the label statements’ materiality: 71% of respondents to a JPMS survey said that a statement that a product was made with “cruelty-free ingredients” was a “most important callout.” And internal JPMS documents also show that emphasizing the cruelty-free nature of JPMS’s products was a key marketing strategy for the firm. This is sufficient evidence that the alleged misrepresentations would be a “substantial factor in” a “large percentage of consumers’ purchasing decisions,” and thus to justify an inference of classwide reliance. See Lytle v. Nutramax Laboratories, Inc., 114 F.4th 1011, 1034– 36 (9th Cir. 2024). Third, the plaintiffs’ counsel are adequate to represent the class. Fourth, the evidence does not show that “no animal testing” is so susceptible of different understandings as to defeat predominance. That term has a sufficiently fixed meaning to enable a classwide determination as to whether putting that term on a label would be misleading to a reasonable consumer if JPMS did in fact arrange for that product or other products to be animal tested in China. Cf., e.g., Jones v. ConAgra Foods, Inc., 2014 WL 2702726, at *14-15 (N.D. Cal. June 13, 2014) (holding otherwise with regard to term “all natural”). JPMS’s presentation to the contrary is quite weak. For one thing, JPMS’s expert didn’t ask consumers what they took the statement to mean—or, more precisely, whether they would take it to represent that a product on which it appeared wasn’t tested on animals in the way the plaintiffs allege JPMS’s products were, or to represent that the company selling that product didn’t test other products that way. The expert asked whether consumers would find it “acceptable” for a company or product using that statement to test on animals in certain ways. This is a different question: Someone who is staunchly opposed to animal testing would say that every form of animal testing is unacceptable no matter the representations by a label or company, while someone with no problems with animal testing would say the opposite. So this analysis does very little to show that consumers’ understandings of “no animal testing” actually vary widely. Moreover, some of the ways that JPMS says consumers would interpret “no animal testing” are simply implausible, or appear to be based on distortions of the record. For instance, JPMS cites the complaint’s allegation that animals are killed after testing is complete to suggest that one might interpret “no animal testing” as meaning that “testing is okay, but animals cannot be killed when testing is complete so a necropsy can be done to determine internal damage.” It is inconceivable that someone would interpret “no animal testing” to mean “we animal test, but we don’t kill the animals afterward.” As another example, JPMS implies that plaintiff Rica Guerrero testified that she understood “no animal testing” to mean that a product bearing that label “contains ‘less chemical or any type of filler ingredients.’” But read in context, Guerrero was obviously stating that the reason she cared about the no animal testing label was that she expected that products not tested on animals would necessarily have fewer chemicals, not that she understood the label to say that was the case. Finally, the plaintiffs have established that their damages can be measured on a classwide basis. As an initial matter, some of the asserted problems with their damages models are taken care of by the fact that, with Heagney as the only class representative, only damages attributable to the “no animal testing” and “JPMS does not conduct or endorse animal testing” statements might need to be measured on a classwide basis.
Free access — add to your briefcase to read the full text and ask questions with AI
UNITED STATES DISTRICT COURT NORTHERN DISTRICT OF CALIFORNIA
RANDALL HEAGNEY, et al., Case No. 23-cv-00687-VC
Plaintiffs, ORDER ON MOTION FOR CLASS v. CERTIFICATION, MOTION TO EXCLUDE, AND MOTION FOR JOHN PAUL MITCHELL SYSTEMS, LEAVE TO AMEND Defendant. Re: Dkt. Nos. 120, 121, 124
The motion for class certification is granted, although with respect to a narrower class than proposed by the plaintiffs. JPMS’s motion to exclude the plaintiffs’ expert witness and the plaintiffs’ motion for leave to amend are denied. This order assumes the reader’s familiarity with the facts, governing legal standards, and arguments made by the parties. Class certification. A class of all California residents who between May 1, 2015, and January 1, 2020, purchased, directly from JPMS or through an authorized third-party retailer or salon, a JPMS hair care product branded as Paul Mitchell, Clean Beauty, Tea Tree, MITCH, Awapuhi Wild Ginger, Neuro, or MVRCK, is certified. Plaintiff Randall Heagney is appointed class representative. Material differences in state laws preclude certification of a nationwide class. With respect to the plaintiffs’ UCC claims, “there are material conflicts between California warranty law and the warranty law of the other forty-nine states.” In re Hitachi TV Optical Block Cases, 2011 WL 9403, at *6 (S.D. Cal. Jan. 3, 2011). There are also differences between California’s consumer protection statutes and those of other states, including the states the plaintiffs (other than Heagney) live in: Virginia, for instance, does not allow class actions, while Texas requires pre-suit notice to the defendant. See Gustafson v. BAC Home Loans Servicing, LP, 294 F.R.D. 529, 539 (C.D. Cal. 2013); Darisse v. Nest Labs, Inc., 2016 WL 4385849, at *10 (N.D. Cal. Aug. 15, 2016). Because the relevant state laws differ materially, classmembers’ claims must be governed by the laws of their home states. See Mazza v. American Honda Motor Co., 666 F.3d 581, 593–94 (9th Cir. 2012). Abstract “reliance” or “non-reliance” subclasses cannot be certified because “[c]hoice of law requires determining the single state’s substantive law that applies to a particular plaintiff’s claim.” Cimoli v. Alacer Corp., 587 F. Supp. 3d 978, 988 (N.D. Cal. 2022) (emphasis added). Nor can subclasses be certified for the states where the other plaintiffs live because the plaintiffs did not bring claims under the laws of those states—and as discussed below, their motion for leave to amend is denied. Because Heagney is the only named plaintiff to whom California law applies, the class claims can only proceed as to the representation he saw and says he relied on—that is, the “no animal testing” statement, as well as the substantially similar “JPMS does not conduct or endorse animal testing” statement. Cf. Bazarganfard v. Club 360 LLC, 344 F.R.D. 411, 426 & n.14 (C.D. Cal. 2023). There is no evidence, and the plaintiffs do not allege, that he saw the “pioneer in cruelty-free haircare” representation. Moreover, he stopped using JPMS products “around 2020,” and that slogan wasn’t used until that year. While it’s a close call based on the evidence the plaintiffs present about industry usage and consumers’ understanding of the terms, that statement is too different from “no animal testing” to be considered substantially the same representation. Cf. Conde v. Sensa, 2019 WL 4277414, at *8 (S.D. Cal. Sep. 10, 2019). For that reason, and because the plaintiffs have not shown how it could be determined which representation was seen by class members who purchased the products during the period that both statements were in the market, it is appropriate for the class period to end in January 2020, when the “pioneer” statement began to be used. The rest of JPMS’s challenges to certification fail. First, although there isn’t adequate proof of classwide exposure to the marketing statements, there is evidence that the targeted representations were visible enough and “consistently made on a product’s label.” Krommenhock v. Post Foods, LLC, 334 F.R.D. 552, 565 (N.D. Cal. 2020). Second, reliance can be inferred on a classwide basis because there is sufficient evidence of the label statements’ materiality: 71% of respondents to a JPMS survey said that a statement that a product was made with “cruelty-free ingredients” was a “most important callout.” And internal JPMS documents also show that emphasizing the cruelty-free nature of JPMS’s products was a key marketing strategy for the firm. This is sufficient evidence that the alleged misrepresentations would be a “substantial factor in” a “large percentage of consumers’ purchasing decisions,” and thus to justify an inference of classwide reliance. See Lytle v. Nutramax Laboratories, Inc., 114 F.4th 1011, 1034– 36 (9th Cir. 2024). Third, the plaintiffs’ counsel are adequate to represent the class. Fourth, the evidence does not show that “no animal testing” is so susceptible of different understandings as to defeat predominance. That term has a sufficiently fixed meaning to enable a classwide determination as to whether putting that term on a label would be misleading to a reasonable consumer if JPMS did in fact arrange for that product or other products to be animal tested in China. Cf., e.g., Jones v. ConAgra Foods, Inc., 2014 WL 2702726, at *14-15 (N.D. Cal. June 13, 2014) (holding otherwise with regard to term “all natural”). JPMS’s presentation to the contrary is quite weak. For one thing, JPMS’s expert didn’t ask consumers what they took the statement to mean—or, more precisely, whether they would take it to represent that a product on which it appeared wasn’t tested on animals in the way the plaintiffs allege JPMS’s products were, or to represent that the company selling that product didn’t test other products that way. The expert asked whether consumers would find it “acceptable” for a company or product using that statement to test on animals in certain ways. This is a different question: Someone who is staunchly opposed to animal testing would say that every form of animal testing is unacceptable no matter the representations by a label or company, while someone with no problems with animal testing would say the opposite. So this analysis does very little to show that consumers’ understandings of “no animal testing” actually vary widely. Moreover, some of the ways that JPMS says consumers would interpret “no animal testing” are simply implausible, or appear to be based on distortions of the record. For instance, JPMS cites the complaint’s allegation that animals are killed after testing is complete to suggest that one might interpret “no animal testing” as meaning that “testing is okay, but animals cannot be killed when testing is complete so a necropsy can be done to determine internal damage.” It is inconceivable that someone would interpret “no animal testing” to mean “we animal test, but we don’t kill the animals afterward.” As another example, JPMS implies that plaintiff Rica Guerrero testified that she understood “no animal testing” to mean that a product bearing that label “contains ‘less chemical or any type of filler ingredients.’” But read in context, Guerrero was obviously stating that the reason she cared about the no animal testing label was that she expected that products not tested on animals would necessarily have fewer chemicals, not that she understood the label to say that was the case. Finally, the plaintiffs have established that their damages can be measured on a classwide basis. As an initial matter, some of the asserted problems with their damages models are taken care of by the fact that, with Heagney as the only class representative, only damages attributable to the “no animal testing” and “JPMS does not conduct or endorse animal testing” statements might need to be measured on a classwide basis. Nor are JPMS’s remaining arguments persuasive. Although the plaintiffs’ expert, Gareth Macartney, has not fully developed his models, he done more than “merely gestur[e] at a model or describ[e] a general method.” Lytle, 114 F.4th at 1032. He has identified certain variables that might be necessary to control for— including other label statements, brand and product characteristics, and marketing channels—and explained how he would determine which variables must be controlled for and how he would control for them in each of his two proposed models. He has also explained how he would get the data necessary and how he would determine the appropriate sampling population for a conjoint survey. In other words, as in Lytle, he may not have “determined the precise demographic makeup of the individuals to be surveyed” and “selected all of the parameters for his model[s],” and may “lack[] certain data needed to finalize his calculations.” Id. at 1031. But he has “chart[ed] a path to obtain[ing] all necessary data and demonstrate[d] that” his proposed method “will be viable” as applied to this case. Jd. at 1032. And he has sufficiently planned to account for supply-side factors because he will use actual market prices and sales data. See Hadley v. Kellogg Sales Co., 324 F. Supp. 3d 1084, 1105-06 (N.D. Cal. 2018). Leave to amend. The motion for leave to amend is denied. Allowing the plaintiffs to amend would substantially prejudice JPMS. As the foregoing discussion indicates, a major issue relevant to whether this case is appropriate for class treatment is the different label statements on JPMS’s products. And as discussed, the statements appearing on those products changed over time. So to extend the class period by 17 years would require additional discovery into previous statements and force JPMS to go through another round of class certification briefing. Adding claims under the laws of three additional states would similarly unfairly require JPMS to litigate another class certification motion, because it would present new issues regarding the manageability of a trial involving claims under the laws of four different states. Finally, the plaintiffs sought leave to amend three months after they acquired the ostensibly new information justifying amendment, on the same day they moved for class certification. This was an unnecessary delay. Excluding the plaintiffs’ expert. JPMS’s motion to exclude Macartney’s testimony is denied. JPMS’s arguments in support of this motion substantially overlap with its arguments as to why the plaintiffs did not establish that damages can be measured on a classwide basis, and those overlapping arguments fail here for the same reasons. Beyond that, JPMS’s objection to Macartney’s literature review fail because a damages expert can assume the merits of a case. Rushing v. Williams-Sonoma, Inc., 2024 WL 779601, at *17 (N.D. Cal. Feb. 21, 2024). IT IS SO ORDERED. Dated: June 27, 2025 Koo. VINCE CHHABRIA United States District Judge