JUUL LABS, INC. v. 4X PODS

Opinion

UNITED STATES DISTRICT COURT FOR THE DISTRICT OF NEW JERSEY

JUUL LABS, INC., Plaintiff, v. Civ. No. 18-15444 (KM) (MAH) 4X PODS, EONSMOKE, LLC d/b/a OPINION 4X PODS, GREGORY GRISHAYEV, MICHAEL TOLMACH, and JOHN DOES 1–50, Defendants.

KEVIN MCNULTY, U.S.D.J.: Juul Labs, Inc. makes e-cigarettes, one component of which is an insertable pod of nicotine and flavoring. Eonsmoke, LLC1 also makes pods and has marketed them with the word “Juul” in social media posts and advertising, while at least sometimes using packaging similar to Juul’s. Juul sued Eonsmoke, alleging, among other things, trademark and trade dress infringement. In response, Eonsmoke began moving funds to avoid having to pay Juul any judgment. To freeze Eonsmoke’s assets, Juul moved for injunctive relief, namely an asset freeze in anticipation of the equitable remedy of disgorgement. I denied the motion, concluding that some limited claims had a likelihood of success on the merits, that it was too early to assess others, and that an asset freeze was too broad a remedy. Juul Labs, Inc. v. 4X PODS, 439 F. Supp. 3d 341, 360–61 (D.N.J. 2020), appeal dismissed, 2020 WL 5240430 (3d Cir. July 24, 2020) (“Juul I”). Following an amendment to the complaint, more discovery, and further development of the factual record, Juul renews its motion. (DE 232.)2 For the following reasons, the motion is GRANTED.

1 I refer to Defendants collectively as “Eonsmoke.” 2 Certain citations to the record are abbreviated as follows: I. BACKGROUND3 A. Facts Juul developed an e-cigarette device and now dominates that market. Juul I, 439 F. Supp. 3d at 345. One component of that device is a pod filled with a proprietary blend of, among other things, liquid nicotine and flavoring. Id. A user inserts the pod into the device (which resembles a USB stick) and inhales. The device then vaporizes the liquid in the pod, allowing the user to “smoke” or “puff” the vapor (hence the term “vaping”). (Am. Compl. ¶ 2; see generally Food & Drug Admin., “Vaporizers, E-Cigarettes, and other Electronic Nicotine Delivery Systems (ENDS)” (Sept. 17, 2020), https://www.fda.gov/tobacco-products/products-ingredients- components/vaporizers-e-cigarettes-and-other-electronic-nicotine-delivery- systems-ends.) Juul makes pods, which it sells either as a component of Juul kits that include a device, or separately. (Thomas Rep. at 7.) Juul trademarked the word “Juul” and its logo and uses a distinctive packaging with those trademarks. (Am. Compl. ¶¶ 24, 26, 28, 30, Ex. 1, 2.) Eonsmoke, an e-cigarette company ran by Gregory Grishayev and Michael Tolmach, developed its own pods, which are compatible with Juul

DE = docket entry Am. Compl. = Amended Complaint (DE 187) Juul Brf. = Brief in Support of Juul’s Renewed Application for a Temporary Restraining Order and an Order to Show Cause Why the Court Should Not Issue a Preliminary Injunction (DE 232) Eonsmoke Opp. = Defendants’ Opposition to Plaintiff’s Renewed Motion for a Preliminary Injunction (DE 238) Tolmach Tr. = Deposition of Michael Tolmach (DE 137-1, Ex. 2) Grishayev Tr. = Deposition of Gregory Grishayev (DE 137-1, Ex. 1) Thomas Rep. = Expert Report of Vincent A. Thomas, CPA, CVA, CFF, ABV (DE 232-3, Ex. 4) Pittaoulis Rep. = Expert Report of Dr. Melissa Pittaoulis, PhD (DE 232-3, Ex. 3) 3 A more exhaustive background can be found in my previous opinion. Here, I set forth facts that have arisen since and the most important facts to the current issues. devices.4 To market those pods, Eonsmoke relied mostly on social media like Instagram, Twitter, Facebook, and Tumblr. (Tolmach Tr. at 74:6–8, 166:1–6, 184:1–7; Grishayev Tr. at 283:16–23.) Eonsmoke’s posts mostly used the same format: (1) an image of the Eonsmoke product, which often included a label that the pod was “Juul compatible,” accompanied by (2) a short caption describing the product or inviting the viewer to purchase it, followed by (3) hashtags. (Thomas Rep. at 14–21.) A hashtag consists of the pound/number symbol (#), followed by text. When a social media user adds a hashtag to a post, the hashtag, which is hyperlinked, acts as a tag. This tagging has two consequences: First, if a viewer of the post clicks the hashtag, the social media platform will take the viewer to a page containing any other posts with that hashtag. Second, if a user searches that hashtag through the platform’s search engine, the platform will take the user to the page containing any posts with that hashtag. Hashtags thus have an indexing or cataloguing function, “allow[ing] people to easily follow topics they are interested in” and “discover content and accounts based on [their] interests.” Twitter, “How to use hashtags,” https://help.twitter.com/en/using- twitter/how-to-use-hashtags (last visited Dec. 12, 2020). In other words, hashtags provide a way to link individual posts to larger topics and conversations. When Eonsmoke entered the market, it included hashtags of other e- cigarette brands in its posts. It used “Juul” the most. Tolmach and Grishayev explained that they used “Juul” in hashtags to “promote” Eonsmoke. (E.g., Tolmach Tr. at 74:6–8; Grishayev Tr. at 283:16–23.) Indeed, Eonsmoke recognized that, given the function of a hashtag, the hashtags would allow Eonsmoke’s posts to be found via the Juul name. (Tolmach Tr. at 214:8–22.) To illustrate the form of Eonsmoke’s social media posts, I reproduce two representative examples, one from Instagram and one from Facebook, which I

4 Eonsmoke also created a subsidiary brand called “4X Pods.” The 4X products and marketing are encompassed by this discussion. will call Post 1 (DE 231-2, Ex. 14, at 3 (sour berry flavor)) and Post 2 (Thomas Rep. at 19 (grape flavor)): : te 4zpod + Fallow 4X pops” 4epod Sour berry palts4| 65mg nic 6.5%! . depod 6 ®8juul #juulcentra! *juulpang Sour Berry Belts Thuleads 65% St 1h farthekid Of Fim for sure copping me a pack Joel Compatitie i pia _ ie eye

Boge WARNING: cs This product ; contains : Z nicotine. x Fae Bh Nicotine is ce an addictive OQ chemical. 45 hes

EonSmoke Electronic Cigarette duly 11, 2018 - We think we are launching Black Cherry next in pods! Yay or Nay? In the meantime here is worlds only Grape 6% for you. #eonsmoke #uul #juuling #juulgang #uulvapor #uulmemes #uulnation #phix #sourin #smok #smoktech #sourin #suorin

& eonsmok FOUR PODS GRAPE ~~ aa

WARNING: This product contains nicotine. Nicotine is an addictive chemical. —————

Not long after Eonsmoke began this marketing strategy, its revenues shot up—from $2.3 million in 2017, to $30 million in 2018, to $90 million in 2019. (Thomas Rep. at 26.) Also during this time, Eonsmoke developed packaging that resembled Juul’s and had retailers display the Juul and Eonsmoke products near one another. Juul I, 439 F. Supp. 3d at 346. B. Juul I Juul sued Eonsmoke. Juul’s original complaint asserted claims for, among other things, trademark infringement of the Juul logo under 15 U.S.C. § 1114 (Count 1) and false designation of origin under § 1125(a) based on infringement of the Juul logo trademark and packaging trade dress (Count 2). (DE 1 ¶¶ 57–72.) Discovery revealed communications between Tolmach and Grishayev “express[ing] in no uncertain terms that they will never pay any judgment that may ultimately be entered in this action.” Juul I, 439 F. Supp. 3d at 347. Juul also found out that Tolmach and Grishayev jointly opened a brokerage account, which they called the “family office account,” and had moved funds from the Eonsmoke business checking account to that account. Id. at 348. Concerned that Eonsmoke was dissipating its assets to avoid paying any judgment, Juul moved for injunctive relief to freeze Eonsmoke’s assets, estimated to be in the tens of millions of dollars. Id. at 349. Before I ruled, Juul moved to amend its complaint. The (then-proposed) Amended Complaint, among other things, added a more specific claim that Eonsmoke committed trademark infringement by using the Juul wordmark in social media posts. (Am. Compl. ¶¶ 85–94.) I held a hearing on Juul’s motion for an injunction in January 2020. I issued a decision in Juul I the next month, holding as follows: • Juul was likely to succeed on its trademark infringement and trade dress claims based on the logo and packaging. 439 F. Supp. 3d at 356, 360. • Irreparable harm was not established because (1) there was not proof that Eonsmoke was dissipating—as opposed to just moving—assets; and (2) few Eonsmoke products with Juul-like packaging or using the Juul logo were sold, with resulting estimated profits of less than $500,000, so freezing tens of millions of dollars in assets was not warranted. Id. at 358–59. • The proposed wordmark claim could potentially recover profits on all Eonsmoke products sold, but Juul had not yet actually asserted or produced evidence on the merits of that claim or the resulting profits. Id. at 358, 360. • Although an asset freeze was not yet warranted, Eonsmoke was to report its finances quarterly to the Court. Id. at 360–61. C. New Developments Since Juul I, some new facts have developed that are the subject of Juul’s renewed motion for an asset freeze. They can be summarized as follows: First, Judge Hammer granted Juul’s motion to amend, so the wordmark claim is now a part of the case. (DE 185.) Second, more assets have been expended or moved. In particular, Tolmach purchased a $2.3 million property in Los Angeles, for which he spent $250,000 on renovations/construction and $91,000 on furniture. (DE 223-1; 227-6; 231-2, Ex. 11, 12.) Next, there was an unexplained $50,000 transfer to an unidentified account. (DE 227-2.) Less significantly, there has been some unexplained smaller-scale consumption, like a $6,500 credit card payment. (DE 227-1.) Third, Tolmach and Grishayev opened new accounts with a different bank, which they call “substitute accounts,” and moved around 90% of assets from the “family office” account to the new accounts. (DE 167-2, at 16–17; 181; 223.) (These movements occurred in January 2020, prior to the issuance of my decision in Juul I, but Eonsmoke did not reveal these movements until after that decision.) Eonsmoke explains that the bank for the previous accounts changed its management and investment policies; the new accounts, it says, were opened simply because they provided “more advantageous account management and investment income.” (DE 167, at 1.) It is important to note that the defense has at all times conceded that these are Eonsmoke’s funds; indeed it has insisted that they are Eonsmoke’s funds in connection with arguing that they have not been dissipated, arguing that any withdrawals for the benefit of the individual defendants are simply routine distributions of profits. Fourth, discovery has continued. Of note are two expert reports submitted by Juul. One is a survey of e-cigarette users conducted by Dr. Melissa Pittaoulis, PhD, who specializes in surveying and sampling, to test whether social media-literate consumers were confused by Eonsmoke’s social media posts as to the origin of their products. (Pittaoulis Rep. at 3, 13.) The second is a report by Vincent A. Thomas, an accountant, which relied on financial data produced by Eonsmoke to estimate Eonsmoke’s sales and profits. (Thomas Rep. at 1–2.) Fifth, Eonsmoke is no longer in business, as of April 2020. (DE 231-1; 182-1 ¶ 6; Tolmach Tr. at 305:25–06:4.) Because of this developing evidence on Juul’s wordmark claim, and because of Eonsmoke’s dwindling assets, Juul renews its motion for an asset freeze. (Juul Brf. at 1.) On December 7, 2020, I held a hearing by video (DE 246), and I have considered the entire record to date. II. DISCUSSION I initially dealt with Juul’s request for an asset freeze in Juul I. Then, as now, Juul sought a freeze of assets to preserve them in anticipation of a judgment granting the equitable remedy of disgorgement of profits.5 This

5 As discussed in more detail in Juul I, the application is therefore not barred by Grupo Mexicano de Desarrollo, S.A. v. Alliance Bond Fund, Inc., 527 U.S. 308 (1999) (barring restraint to preserve assets in anticipation of award of damages at law, but distinguishing the case of equitable remedies). See Sweet People Apparel, Inc. v. Fame of NY, Inc., Civ. No. 11-1666, 2011 WL 2937360 at * (D.N.J. July 19, 2011) (Grupo Mexicano permits prejudgment asset freeze in case seeking equitable accounting of defendants’ profits under the Copyright and Trademark Acts). opinion assumes familiarity with Juul I, which is incorporated by reference, and is necessary background for understanding this opinion, which is supplemental. In Juul I, I held that Juul had not shown that it was then entitled to an asset freeze, but I previewed what was missing from its application: Because the packaging-related trademark claims were worth no more than $500,000, Juul would need to show that (1) it could succeed on its additional wordmark claim, (2) profits flowing from wordmark infringement were of such an amount to support an asset freeze, and (3) Eonsmoke’s conduct warranted an asset freeze. 439 F. Supp. 3d at 360. The issue now is whether the amended claims and facts developed in discovery support those necessary showings. A. Standard of Review In deciding whether to grant a preliminary injunction, I consider: (1) the likelihood that the plaintiff will prevail on the merits at final hearing; (2) the extent to which the plaintiff is being irreparably harmed by the conduct complained of; (3) the extent to which the defendant will suffer irreparable harm if the preliminary injunction is issued; and (4) that the public interest weighs in favor of granting the injunction. Ramsay v. Nat’l Bd. of Med. Exam’rs, 968 F.3d 251, 246 (3d Cir. 2020) (citation and alterations omitted). The first two factors are essential. Reilly v. City of Harrisburg, 858 F.3d 173, 179 (3d Cir. 2017). B. Likelihood of Success on the Merits of the Wordmark Claim Juul must show “a reasonable chance or probability[] of winning” its wordmark claim. Ramsay, 968 F.3d at 256 (citation omitted). This is not to imply, however, that Juul must meet its ultimate burden of “a more-likely- than-not showing of success.” Reilly, 858 F.3d at 179 n.3. To succeed on its wordmark claim, Juul must show that “(1) it has a valid and legally protectable mark; (2) it owns the mark; and (3) [Eonsmoke’s] use of the mark to identify goods or services causes a likelihood of confusion.” A&H Sportswear Inc. v. Victoria’s Secret Stores Inc., 237 F.3d 198, 210 (3d Cir. 2000). Juul already has made a showing on the first two elements, and they are not in dispute. Juul I, 439 F. Supp. 3d at 356. The key issue is the third element: whether Eonsmoke’s use of the wordmark “Juul” in its social media and online advertising creates a likelihood of confusion. Courts in this Circuit usually determine likelihood of confusion by weighing the ten “Lapp factors.” Freedom Card, Inc. v. JPMorgan Chase & Co., 432 F.3d 463, 471 (3d Cir. 2005) (citing Interpace Corp. v. Lapp, Inc., 721 F.2d 460, 463 (3d Cir. 1983)).6 The analysis differs, however, in cases like this one, that involve nominative fair use. Nominative fair use occurs when the defendant uses the plaintiff’s trademark to describe the defendant’s own product, not to imply that the product originates with the plaintiff. Century 21 Real Estate Corp. v. Lendingtree, Inc., 425 F.3d 211, 214 (3d Cir. 2005). An example would be an assurance of compatibility, i.e., when Company A uses Company B’s wordmark to advertise that Company A’s products are compatible with Company B’s. Keurig, Inc. v. Strum Foods, Inc., 769 F. Supp. 2d 699, 704, 707–08 (D. Del. 2011) (maker of coffee cartridges stated on its packaging that the cartridges

6 Those factors are (1) the degree of similarity between the owner’s mark and the alleged infringing mark; (2) the strength of the owner's mark; (3) the price of the goods and other factors indicative of the care and attention expected of consumers when making a purchase; (4) the length of time the defendant has used the mark without evidence of actual confusion arising; (5) the intent of the defendant in adopting the mark; (6) the evidence of actual confusion; (7) whether the goods, though not competing, are marketed through the same channels of trade and advertised through the same media; (8) the extent to which the targets of the parties’ sales efforts are the same; (9) the relationship of the goods in the minds of consumers because of the similarity of function; (10) other factors suggesting that the consuming public might expect the prior owner to manufacture a product in the defendant's market, or that he is likely to expand into that market. Freedom Card, 432 F.3d at 471 (quoting Lapp, 721 F.2d at 463). were “[f]or use by owners of Keurig® coffee makers”). In such a case, the Lapp analysis is modified in two ways. First, some Lapp factors become close to irrelevant, while others take on greater importance. Century 21, 425 F.3d at 224–25. The first two factors, similarity of the marks and strength of the owner’s mark, do not receive weight. It is inherent in fair use that the defendant used the plaintiff’s mark and relied on the strength thereof (although only so far as principles of fair use permit). Id. Which factors instead come to the forefront, Century 21 explained, is left to the district court. Id. at 226. Nonetheless, Century 21 suggested that four factors will typically be important: (1) evidence of actual confusion, (2) the length of time the defendant used the mark without actual confusion, (3) the defendant’s intent, and (4) the price of the goods and other factors indicative of the care and attention expected of consumers when making a purchase. Id. at 225–26. Next, a second step is added. If the plaintiff can show likelihood of confusion based on the modified Lapp factor test, then the burden shifts to the defendant to show nominative fair use. Id. at 222. A defendant must show (1) that the use of plaintiff’s mark is necessary to describe both the plaintiff’s product or service and the defendant’s product or service; (2) that the defendant uses only so much of the plaintiff’s mark as is necessary to describe plaintiff’s product; and (3) that the defendant’s conduct or language reflect the true and accurate relationship between plaintiff and defendant’s products or services. Id. (formatting altered). Eonsmoke, in its briefs and at oral argument, seems to be arguing that it made nominative fair use of the Juul wordmark as an assurance that its own product was compatible with Juul e-cigarettes. (See, e.g., DE 130 at 10.) Although nominative fair use is an affirmative defense, Century 21, 425 F.3d at 228, and Eonsmoke’s answers have not pleaded the defense explicitly (DE 215), I will apply Century 21’s bifurcated test for three reasons: (1) Eonsmoke has raised the issue, albeit just barely, and Juul has responded accordingly (plus discovery has focused on compatibility), see In re Frescati Shipping Co., 886 F.3d 291, 313 (3d Cir. 2018) (affirmative defenses can be considered even if less than clearly raised, provided there is no prejudice to the other party), aff’d sub nom. CITGO Asphalt Refining Co. v. Frescati Shipping Co., 140 S. Ct. 1081 (2020); (2) the typical Lapp analysis is not ideally suited to addressing arguments relating to nominative fair use, Century 21, 425 F.3d at 224; and (3) in my equitable discretion, it would be unfair to take the extraordinary step of freezing Eonsmoke’s assets without giving full consideration to the most relevant defense, see eBay, Inc. v. MercExchange, LLC, 547 U.S. 388, 391 (2006) (principles of equity apply to a request for injunctive relief and the decision is left to a court’s equitable discretion). I turn to the applicable factors. Likelihood of Confusion I highlight the four Lapp factors which Century 21 identified as important and then consider any impact of the remaining factors. i. Actual Confusion Evidence of actual confusion gets to “the heart of the nominative fair use situation” here. Century 21, 425 F.3d at 226. The question is whether a consumer would think that Eonsmoke’s pods are associated with Juul, or would instead realize that Eonsmoke’s pods are distinct products that are merely compatible with Juul devices. Id. Indeed, in any trademark case, “evidence of actual confusion may be highly probative.” Checkpoint Sys., Inc. v. Check Point Software Techs., Inc., 269 F.3d 270, 291 (3d Cir. 2001). Such evidence may take the form of consumer surveys or anecdotal evidence. See Sabinsa Corp. v. Creative Compounds, LLC, 609 F.3d 175, 187 (3d Cir. 2010). On this motion, Juul has submitted a consumer survey conducted by Dr. Pittaoulis. (Pittaoulis Rep. at 3, 13.) A test group was shown an Eonsmoke social media post with Juul hashtags and “JUUL COMPATIBLE” both in the caption and in the image of an Eonsmoke package. A control group was shown the same post, but with the Juul hashtags changed to Eonsmoke hashtags and “JUUL COMPATIBLE” changed to “COMPATIBLE PODS.” (Id. at 10.) Respondents then answered questions probing their beliefs as to who they thought put out the posts and sold the products. (Id. at 18–20.) Approximately 56.8% of the test group versus 10.6% of the control group thought that the post was either issued by Juul, affiliated with Juul, or authorized by Juul, yielding a net confusion rate of 46.2% (the test group minus the control group). (Id. at 10.) I give this evidence substantial weight. The 46.2% confusion rate well exceeds the 15% rate which the Third Circuit has found “sufficient to demonstrate actual confusion.” Novartis Consumer Health, Inc. v. Johnson & Johnson-Merck Consumer Pharms. Co., 290 F.3d 578, 594 (3d Cir. 2002); see also J. Thomas McCarthy, 6 McCarthy on Trademarks and Unfair Competition § 32:188 (5th ed. 2020) (collecting cases showing that “figures in the range of 25% to 50% have been viewed as solid support for a finding of a likelihood of confusion”). The survey methodology can be criticized in that it does not test all variables, one variable at a time. By hypothesis, Eonsmoke could include the statement “Juul Compatible” as a legitimate compatibility assurance. See Keurig, 769 F. Supp. 2d at 707–08. An underlying issue, then, is whether the additional uses of the Juul wordmark (hashtags, in particular) exceeded the bounds of a compatibility assurance and created an unacceptable level of consumer confusion. So the designer of the survey might profitably have considered holding at least one compatibility assurance statement constant as between the control and test group, changing only the hashtags. See THOIP v. Walt Disney Co., 690 F. Supp. 2d 218, 240 (S.D.N.Y. 2010) (“To fulfill its function, a control should share as many characteristics with the experimental stimulus as possible, with the key exception of the characteristic whose influence is being assessed.” (internal quotation marks and citation omitted)). To say that the survey could have been supplemented, however, is not to say that it lacks value. The survey at least shows that (1) actual Eonsmoke posts confused 56.8% of viewers, and (2) the presence of Juul hashtags likely played some role in that confusion. Accordingly, at this preliminary stage, the survey constitutes substantial evidence in favor of Juul.7 Surveys are only circumstantial evidence of actual confusion; they involve “an experimental environment,” not “real consumers making mistaken purchases.” McCarthy on Trademarks § 32:184. Anecdotal evidence can be more direct evidence of actual confusion, id., and so is “both relevant and probative,” Freedom Card, 432 F.3d at 478. Juul produces some anecdotal evidence of actual confusion on the part of actual consumers of Eonsmoke products, but it is not particularly strong. First, Eonsmoke’s general manager testified that she had received calls asking for Juul pods. (DE 231-2, Ex. 16 at 108:6–109:3.) Next, Juul points to a comment on an Eonsmoke Facebook post stating “all i want is a juul eonsmoke pod kit and those capsules for my vape pen.” (Ex. 13 at 2.)8 The reference to “juul eonsmoke pod kit” could show that the user conceptually merges Juul and Eonsmoke, or does not care. The statement remains open to interpretation, and we know nothing about the commenter. I therefore give some, but not

7 Eonsmoke asks me to ignore Juul’s expert reports because (1) Eonsmoke’s rebuttal expert reports are not yet due under the Magistrate Judge’s discovery schedule, and (2) Eonsmoke will make a Daubert motion to exclude Juul’s reports. (Eonsmoke Opp. at 3.) I take Eonsmoke’s point that Juul’s expert reports may well be rebutted or diminished in significance as the case progresses. Still, this is a preliminary injunction, not summary judgment or trial. My task at the preliminary injunction stage is to determine whether Juul has produced enough evidence to show a reasonable probability of succeeding on the merits. See St. Thomas-St. John Hotel & Tourism Ass’n v. Gov’t of U.S.V.I., 357 F.3d 297, 301 (3d Cir. 2004). What I have before me is a probative, if imperfect, expert report, and little in opposition but a statement of counsel that it can or will be rebutted. Discovery deadlines aside, nothing prevented Eonsmoke from submitting fact or expert evidence in opposition (or even just argument) to the preliminary injunction application. 8 Juul alleges that it would have more evidence of this type but for Eonsmoke’s spoliation of most social media documents. (Juul Brf. at 11 n.7) Accordingly, Juul has moved separately for an adverse inference as a sanction. (Id.) I do not complicate this decision with a ruling as to the adverse inference because Juul’s evidence on the actual confusion factor is adequate without it. much, weight to this anecdotal evidence. Checkpoint, 269 F.3d at 298–99. Standing alone, it demonstrates only “isolated instances” of actual confusion. A&H Sportswear, 237 F.3d at 227 (citations omitted); see also Harp v. Rahme, 984 F. Supp. 2d 398, 417 (E.D. Pa. 2013) (explaining that the inquiry should focus on whether there is a pattern of confusion, not just one-off examples). All in all, while direct evidence is scant, I must consider it in the context of the consumer surveys, which are robust and significant. Juul has made a fair showing on the actual confusion factor. Because this factor is important in a compatibility assurance case, and because such evidence is usually hard to come by, Juul’s showing weighs heavily in my decision. ii. Length of Time of Actual Confusion The next pertinent Lapp factor asks whether Eonsmoke used the Juul wordmark for a long time without evidence of actual confusion. Century 21, 425 F.3d at 226. Where a mark has been used for a long time with few instances of actual confusion, a court is less likely to find likelihood of confusion. Checkpoint, 269 F.3d at 298–99. There is little evidence on this factor one way or the other. It consists largely of the same evidence offered on the actual confusion factor, supra. Given that both parties seem unable to produce such evidence, it may simply be unavailable, and I will not treat its absence as dispositive. Checkpoint, 269 F.3d at 298–99. This factor favors neither party; more pertinently, it does not significantly advance the case of Juul, which has the burden here. iii. Eonsmoke’s Intent The next pertinent Juul factor concerns the defendant’s intent in using the mark. Discerning the defendant’s intent allows the court to understand whether the defendant’s purpose in using the mark was to “caus[e] consumers to think the plaintiff endorses or sponsors plaintiff’s good or service” and “convey a connection between the parties that may not exist.” Century 21, 425 F.3d at 226. In particular, defendants’ intent may illuminate the issue of whether a supposed assurance of compatibility is actually an effort to exploit a plaintiff’s brand and deceive a consumer as to the source of defendant’s product. See id; Keurig, 769 F. Supp. 2d at 708. Tolmach and Griyashev were questioned about their intent in using Juul’s wordmark, and their answers are revealing. First, Tolmach was evasive in response to certain questions about the hashtags. (Tolmach Tr. at 174:1–5 (“What is the purpose of using #juul with your products? . . . No purpose, no purpose.”).) One can infer from such obstinacy that the true answer is not favorable. See Mali v. Fed. Ins. Co., 720 F.3d 387, 392–93 (2d Cir. 2013) (explaining the concept of permissive adverse inferences, which are not sanctions but the process of inferring one fact from the presence of another). When pressed, Tolmach explained that “I was just throwing [expletive] against the wall to see if it will stick. This is a common strategy in business.” (Tolmach Tr. at 174:14–16.) He also explained that the hashtags were used to “fill up the bulk of the post.” (Id. at 167:5–8.) These are not credible explanations. It strains reason to think that the repeated invocation of Juul’s wordmark on multiple posts was random or just “filler”—or that throwing things against the wall is common business strategy.9 Finally, Tolmach admitted that he used Juul hashtags to “promote” Eonsmoke products (id. at 74:6–8, 166:1–6, 184:1– 7) and that to some extent he used Juul hashtags so that internet searches for “Juul” would lead the searcher to Eonsmoke pages (id. at 214:8–22). Grishayev, too, admitted that the Juul name and hashtags were used to “promote” the product and “get more exposure.” (Grishayev Tr. at 283:16–23.) Such testimony, as might be expected, falls short of a full confession of wrongdoing. It does tend to demonstrate, however, that Eonsmoke’s use of the Juul wordmark exceeded the scope of a typical compatibility assurance. Even interpreted in the light most favorable to Eonsmoke, the testimony suggests

9 Tolmach also explained that he used hashtags with other e-cigarette brand names, too. (E.g., Tolmach Tr. at 170:13–19.) However, several of the posts submitted as evidence used only Juul hashtags or used other brand names to a far lesser extent. The weighting of references in hashtags may be read to reflect Juul’s dominant market share. that Eonsmoke’s purpose went beyond simply indicating to the customer that their pods would work with Juul e-cigarettes (that is, a nominative fair use). Instead, Tolmach and Grishayev both testified that they used the Juul wordmark to steer consumers to their products and promote their products. A key consideration here is whether the defendant used the mark as a way to fairly describe its own product, or whether the description crossed the line to implying an association with the plaintiff’s product or an endorsement from the plaintiff. See Century 21, 425 F.3d at 226; see also Bijur Lubricating Corp. v. Devco Corp., 332 F. Supp. 2d 722, 733 (D.N.J. 2004). Eonsmoke’s avowed purpose in using the hashtags was not merely to convey the information that the pods would fit the same vaping devices as Juul’s pods; it was to piggyback on Juul’s online presence and name recognition. Accordingly, the testimony displaying Eonsmoke’s intent weighs in favor of finding a likelihood of confusion. Eonsmoke’s evident purpose in using the mark was to blur the real relationship (i.e., none) between Juul and Eonsmoke, and to use the Juul name to “promote” Eonsmoke’s products. iv. Consumer Care and Attention The level of care a consumer exercises when purchasing a product is important. If less care is exercised, a consumer can be more easily confused into purchasing an infringer’s product. See Century 21, 425 F.3d at 226. I have already explained that “the relatively inexpensive price tag” for Juul and Eonsmoke products implies that consumers may exercise a relatively low level of care when purchasing those products. Juul I, 439 F. Supp. 3d at 354–55, 357. This is a consumable item; a consumer would not exercise the same caution he or she employs in buying an appliance or an automobile. As to the hashtags, common experience suggests that a typical social media user tends to scroll through posts to glean the essentials, without necessarily poring over the fine print. There is at least a risk that the quick-eyed user, having searched for Juul, would conclude that he or she had found Juul, and perceive some association without drilling down to the issue of mere compatibility. Thus, this factor also weighs in Juul’s favor. v. Other Factors Century 21 left open the issue of whether and to what extent the remaining Lapp factors bear on a nominative fair use case. As the analysis above shows, the four key Lapp factors already favor Juul, so the issue becomes whether the remaining factors would tip the scales back to Eonsmoke. I have already explained that “[t]he seventh, eighth, and ninth factors— the nature of the services provided, the customers targeted, and the methods used to reach those customers—weigh in Juul’s favor” because “the parties operate in the same industries, offer very similar products and target the same audiences, often through social media and online advertising.” Juul I, 439 F. Supp. 3d at 357–58 (internal quotation marks and citation omitted). However, Century 21 suggested that these factors might deserve little weight because, in most cases, there will be substantial overlap between the plaintiff’s and defendant’s businesses; the very purpose of using the mark in a “compatibility” case is so that the relevant audience will recognize it, albeit for a limited, legitimate purpose. 425 F.3d at 226. Thus, weighing these factors would invariably place a thumb on the scale for plaintiffs. See id. Accordingly, I do not give these factors great weight, but in any event, they would not favor Eonsmoke. Likewise, I find little in the tenth factor (other factors suggesting the public might expect the prior owner to manufacture both products) to influence my decision. Plaintiff and defendant are expected to be competitors. “[T]o the extent that the consuming public expects plaintiffs to offer defendant’s services, to offer services in the defendant’s market, or to expand into the defendant's market, this factor has no bearing.” Id. at 249 (Fisher, J., concurring in part, dissenting in part, and concurring in the judgment).10 Juul and Eonsmoke operate in the same market, so such overlaps have no particular bearing. * * * Juul has made a sufficient showing on the four key factors, particularly actual confusion and defendants’ intent. The remaining factors, to the extent they would be considered, do not significantly sway the analysis. Accordingly, Juul has demonstrated a likelihood of confusion, and the inquiry turns to whether Eonsmoke’s use of the Juul wordmark is nevertheless a nominative fair use. Nominative Fair Use Despite the likelihood of confusion, if Eonsmoke’s use of the Juul wordmark qualifies as a nominative fair use, then there is no infringement. This inquiry breaks down to three questions: (1) whether the use of the mark is necessary to describe Juul’s and Eonsmoke’s products, (2) whether only so much of the mark is used as is necessary, and (3) whether an accurate relationship between Eonsmoke and Juul is reflected. Century 21, 425 F.3d at 228. I address each in turn and conclude that none favors Eonsmoke. i. Necessity of Use The first prong asks whether, to fulfill the legitimate fair-use purpose, it is necessary to use the mark. This entails two inquiries: (A) Does Eonsmoke need to use Juul’s mark to assure consumers that its pods are Juul- compatible? (B) Does Eonsmoke need to use Juul’s mark to describe Eonsmoke products? See id. at 229. Were this a case where Eonsmoke simply made a statement that its pods were Juul-compatible, this first prong could be satisfied. See Keurig, 769 F. Supp. 2d at 709. Pods are one component of an e-cigarette. For Eonsmoke to

10 The Century 21 majority did not explain the import of this factor in nominative fair use cases, so I rely on Judge Fisher’s opinion that otherwise agreed with the majority on how the factors applied. get consumers to buy its pods for use in Juul’s e-cigarette, it must inform consumers that the pods are Juul-compatible. Given Juul’s dominance in the market for vaping devices, it probably would not be effective for Eonsmoke to state that its pods are “e-cigarette compatible.” To access the majority of the market, Eonsmoke needs to convey that its products are specifically compatible with Juul’s devices. See Century 21, 425 F.3d at 229 (explaining that a key consideration is whether prohibiting use of the mark would be “a forced reversion to second-best communications”). The problem here is that Eonsmoke did not make a simple compatibility assurance, compare Keurig, 769 F. Supp. 2d at 709, but used “Juul” in hashtags. Now, for obvious reasons, the use of hashtags is not a mode of infringement that has a long history. Still, every court to have considered the use of hashtags with a wordmark has held that such use either supported granting a preliminary injunction or stated an infringement claim. See Museum of Modern Art v. MOMACHA IP LLC, 339 F. Supp. 3d 361, 377–78 (S.D.N.Y. 2018) (preliminary injunction); Pub. Impact, LLC v. Bos. Consulting Grp., Inc., 169 F. Supp. 3d 278, 290 (D. Mass. 2016) (same); Chanel, Inc. v. WGACA, LLC, No. 18 Civ. 2253 (LLS), 2018 WL 4440507, at *2 (S.D.N.Y. Sept. 14, 2018) (motion to dismiss); Fraternity Collection, LLC v. Fargnoli, No. 3:13-CV-664- CWR-FKB, 2015 WL 1486375, at *2 (S.D. Miss. Mar. 31, 2015) (same); Khaled v. Bordenave, 18 Civ. 5187 (PAE), 2019 WL 1894321, at *6 (S.D.N.Y. Apr. 29, 2019) (motion to amend). The case most on point is Align Technologies, Inc. v. Strauss Diamond Instruments, Inc., No. 18-cv-06663-TSH, 2019 WL 1586776, at *1 (N.D. Cal. Apr. 12, 2019). I say that for two reasons: (1) Align is a compatibility assurance case, and (2) the Third Circuit has adapted its version of the nominative fair use test from that of the Ninth Circuit, where Align arose. See Century 21, 425 F.3d at 226 (adopting and modifying the Ninth Circuit’s test). The plaintiff in Align sold the Invisalign orthodontic system and also the iTero scanner, an intraoral scanner used to take images of patients’ teeth to make Invisalign trays. One component of the iTero scanner was a protective sleeve. Defendant Strauss made the “MagicSleeve,” a substitute sleeve for use with the iTero scanner. Strauss marketed the MagicSleeve online, using hashtags containing Align’s “Invisalign” and “iTero” wordmarks. Id. at *2, 6. The U.S. District Court for the Northern District of California held that Strauss’s hashtags did not constitute a nominative fair use of the Invisalign and iTero marks, for a few reasons. First, the hashtags were not necessary to identify MagicSleeve or convey that it was compatible with the iTero scanner. Indeed, the hashtags did not convey that; instead, “[a]ll the reader can glean from the hashtags is an implied association.” Id. Second, because MagicSleeve and Align hashtags were all used together in MagicSleeve posts, a viewer would tend to perceive them all as referring to MagicSleeve, creating an implied association. Id. Third, given the function of a hashtag, Strauss’s posts would come up as a result when someone searched for Align’s marks. Thus, the use of hashtags showed that MagicSleeve sought to accomplish something more than a compatibility assurance to the buyer of the MagicSleeve product. Id. at *7. The same problems are present here. The hashtags do not say that the Eonsmoke pods are Juul-compatible. Nor are they necessary to tell the Eonsmoke customer that the Eonsmoke pod will work with Juul e-cigarettes. A valid compatibility assurance could be accomplished by a simple statement to that effect, and indeed, some of Eonsmoke’s posts and packages contain just such a statement. (E.g., Post 2.) But Eonsmoke goes further and adds multiple Juul hashtags (id.), which primarily serve only to divert customers searching online for Juul and to build a more general association between Juul and Eonsmoke. Align, 2019 WL 1586776, at *6. What is more, some posts do not include a compatibility statement at all, but simply display an image of an Eonsmoke product. (Post 1.) If this is a message of compatibility, it is one in which Eonsmoke seems to be actually relying on an implied association with Juul to convey that message, which is unstated. As in Align, it is hard to credit this as a compatibility assurance when it makes no reference to compatibility. Something else is at work here. The hashtags are not merely means of conveying compatibility information, and they are not “necessary” to market Eonsmoke’s products as Juul-compatible. ii. Degree of Use “[T]he second prong tests only whether the quantum of the plaintiff’s mark used by the defendant was appropriate.” Century 21, 425 F.3d at 230. Three facts weigh against finding in Eonsmoke’s favor on this factor. First, the posts use many Juul hashtags and in greater numbers than other hashtags. Perhaps one Juul hashtag would not leave much of an impression of association between Juul and Eonsmoke, but that is not the case here. Second, the posts use not only “#juul” but also combination forms and variations, like “#juulgang” and “#juulnation” (e.g., Post 2), suggesting that Eonsmoke products are pitched at, or that purchasing them will make the buyer a member of, that Juul “gang” or “nation.” This goes far beyond a simple appending of “#juul” to a post. Third, there are multiple posts using Juul hashtags. (Thomas Rep. at 14–22) This bespeaks a coordinated strategy. One post with Juul hashtags among many unoffending posts could be written off, but that is not the case here. Given how many Juul-related hashtags were used and in how many posts, Eonsmoke stretched its usage of the Juul wordmark beyond nominative fair use. iii. Accurate Presentation The final factor asks whether Eonsmoke has accurately portrayed its relationship with Juul. The court will consider, besides the use of the mark itself, Eonsmoke’s other conduct and language. Century 21, 425 F.3d at 230– 31. As noted, some posts do not include a compatibility statement at all, so a viewer may not perceive, correctly, that Eonsmoke and Juul make separate, compatible products. If anything, a viewer is likely left confused about the contours of Juul and Eonsmoke’s relationship because the posts indiscriminately blend the Juul wordmark with the Eonsmoke brand. (E.g., Post 2.) Indeed, this case is unlike Keurig, which found for the defendant on this prong, because the defendant included only one compatibility statement and a clear disclaimer that the two brands were not affiliated. 769 F. Supp. 2d at 709. Accordingly, the posts do not accurately portray the compatibility relationship between Juul and Eonsmoke. * * * In sum, none of the nominative fair use factors favors Eonsmoke. As a result, Juul’s showing of likelihood of confusion stands unrebutted, so Juul is likely to succeed on its claim that Eonsmoke’s use of the Juul wordmark in social media posts and online advertising is infringing. C. Irreparable Harm: Profits from Wordmark Infringement I next decide whether Juul has “demonstrate[d] potential harm which cannot be redressed by a legal or an equitable remedy following a trial,” and whether that harm is “likely to occur in the absence of an injunction.” Ramsay, 968 F.3d at 262 (quotation marks and citations omitted). The theory of irreparable harm here springs from the Lanham Act’s disgorgement remedy and equitable principles. Lanham Act plaintiffs may recover a “defendant’s profits,” 15 U.S.C. § 1117(a), which Juul seeks (Am. Compl., Prayer D, E). Unlike unliquidated damages at law, those profits would effectively be regarded as Juul’s property, held by Eonsmoke. See Liu v. SEC, 140 S. Ct. 1936, 1944– 45 (2020). If those profits were spent in advance of judgment, then Juul would be irreparably harmed, because it could not recover its property. Gucci Am., Inc. v. Weixing Li, 768 F.3d 122, 132–33 (2d Cir. 2014). Thus, Juul can secure the preliminary equitable remedy of an asset freeze with “[1] a showing that [Juul] [is] likely to become entitled to the encumbered funds upon final judgment and [2] a showing that without the preliminary injunction, [Juul] will probably be unable to recover those funds.” Juul I, 439 F. Supp. 3d at 358 (citation omitted). As previewed in Juul I, those principles as applied here require that Juul must show that the assets it seeks to freeze are linked to the wordmark infringement, and then Juul must show that, absent an injunction, those assets will be diminished or disappear by the time the case comes to final judgment. Id. at 358–59. Link Between the Claims and Assets to Be Frozen Juul is only entitled to an asset freeze of the amount of funds which Juul could reasonably expect to recover via disgorgement based on its wordmark claim. Hoxworth v. Blinder, Robinson & Co., 903 F.2d 186, 198 (3d Cir. 1990) (“[T]he court must make some attempt reasonably to relate the value of the assets encumbered to the likely value of the expected judgment.”); see also Levi Strauss & Co. v. Sunrise Int’l Trading Inc., 51 F.3d 982, 987 (11th Cir. 1995) (district courts in Lanham Act cases have “the authority to freeze those assets which could have been used to satisfy an equitable award of profits”).11 To

11 Before awarding disgorgement, the Third Circuit has traditionally required courts to consider equitable factors, such as (1) whether the defendant had the intent to confuse or deceive, (2) whether sales have been diverted, (3) the adequacy of other remedies, (4) any unreasonable delay by the plaintiff in asserting his rights, (5) the public interest in making the misconduct unprofitable, and (6) whether it is a case of palming off. Banjo Buddies, Inc. v. Renosky, 399 F.3d 168, 175 (3d Cir. 2005) (citation omitted); see also World Ent. Inc. v. Brown, 487 F. App’x 758, 762 (3d Cir. 2012). More recently, however, the Supreme Court has suggested that there should be no “inflexible precondition[s]” to a disgorgement award. Romag Fasteners, Inc. v. Fossil, Inc., 140 S. Ct. 1492, 1497 (2020) (rejecting the view in some Circuits that a finding of willfulness is a prerequisite to a disgorgement award). At this preliminary stage, I need only decide if it is likely that Juul will recover a disgorgement award. Enough of the Banjo Buddies factors, to the extent they endure, would support disgorgement: (1) there is strong evidence of Eonsmoke’s intent to unfairly piggyback off of Juul’s brand, see Section II.B.1.iii, supra; (2) Mr. Thomas’s report estimates significant lost sales (Thomas Rep. at 48); (3) the infringing conduct has ceased, so disgorgement is one of the few remedies left available; (4) Juul has not delayed but tried to quickly move this litigation; (5) the public interest is served by adding this court’s voice to the small chorus of courts recognizing claims of infringement via the still-emerging phenomenon of social media marketing; and (6) using the Juul wordmark without even a modification tends to suggest a form of palming off. See Banjo Buddies, 399 F.3d at 176. calculate disgorgement, the plaintiff first need only estimate the “defendant’s sales,” then the burden shifts to the defendant to deduct costs and show what portion of those sales are not attributable to the infringing conduct. 15 U.S.C. § 1117(a); Mishawaka Rubber & Woolen Mfg. Co. v. S. S. Kresge Co., 316 U.S. 203, 206–07 (1942). Juul submitted an expert report by Mr. Thomas, relying on financial data produced by Eonsmoke. (Thomas Rep. at 1–2.) The report estimates that Eonsmoke’s total sales amounted to $123 million, dating from when Eonsmoke launched its line of pods in 2017 to when it stopped selling pods due to regulatory scrutiny in 2019. (Id. at 11, 25, 56–57.) The report further estimates profits to have been around $49 million, calculated by multiplying the sales for a given year by the profit margin for that year. (The profit margin is taken from Eonsmoke’s profit and loss statements.) (Id. at 5, 56–60.) According to Mr. Thomas, Eonsmoke’s data was incomplete and misleading, so these are conservative estimates. (Id.) For its part, Eonsmoke has not proffered any alternative figure for profits derived from the infringing social media marketing (although Eonsmoke promises that its own to-be-filed expert reports and motions in limine will rebut Juul’s report). (Eonsmoke Opp. at 3.)12

12 The figures I noted at the time of Juul I were lower, but within the same order of magnitude: Juul also highlighted testimony where Mr. Tolmach admitted that he and Mr. Grishayev personally made a total of $40 million selling Juul- compatible pods (DE 137-1 at 23), and that defendants have generated over $70 million in revenue from the sale of Juul-compatible products. (See January 6, 2020 transcript at 48). Juul did not parse out what portion of this $70 million in revenue was attributable to profits from the sale of products that allegedly infringe Juul’s Pod Logo mark or Juul’s packaging trade dress (the claims at issue in the Verified Complaint (DE 1)). Nor did Juul outline what portion of this $70 million was derived from the sale of products that purportedly infringe the Juul wordmark (the new claim added in the proposed Amended Complaint (DE 109-1)). As to the allegedly infringing 4X PODS products, however—the subject of the currently operative complaint—Juul concedes that total profits were approximately $400,000 to $500,000. (January 6, 2020 transcript at 50). Ultimately, when a defendant fails to meet its shifted burden to apportion profits, the court cannot simply base a disgorgement award on gross sales; rather, the court must find support in the record for a reasonable estimation of actual profits from the infringing conduct. Covertech Fabricating Inc. v. TVM Bldg. Prods., Inc., 855 F.3d 163, 177 (3d Cir. 2017). However, at this stage, when the issue is protection of a future judgment, the situation presents different equities. “[A]t a preliminary stage of the litigation,” the Hoxworth court explained, a district court is not required to “make an extended effort to match the value of assets encumbered to the expected value of the judgment dollar for dollar.” 903 F.3d at 199. There must be “some attempt to tailor the scope of the injunction to the likely size of the judgment,” but “the determination regarding how precise that tailoring must or can be in any particular case lies within the equitable discretion of the district court.” Id.; see also Elliott v. Kiesewetter, 98 F.3d 47, 58 (3d Cir. 1996) (affirming an asset freeze consistent with Hoxworth). Thus, while Covertech indicates that final disgorgement awards must have a sufficient basis in evidence, irrespective of a defendant’s failure to meet its shifted burden, Hoxworth acknowledges that the record is more limited at the preliminary stages of litigation, so the estimate required is less exacting and subject to the particular equities of the situation.13 The asset freeze which Juul seeks would be equitably tailored based on this preliminary record. Start with the fact that Eonsmoke has around $20–25 million in cash assets that could be frozen. (Eonsmoke Opp. at 4; DE 227.) Juul estimates total profits to be double that; so if even only half of those profits are attributable to the infringing conduct, then an asset freeze is proportional and tailored. See Neighborhood Assist. Corp. of Am. v. First One

Juul I, 439 F. Supp. at 349. 13 To the extent Hoxworth and Covertech conflict, Hoxworth is both procedurally more on-point and, as the older case, controlling. Holland v. N.J. Dep’t of Corrs., 246 F.3d 267, 278 n.8 (3d Cir. 2001) (“[T]o the extent that [a case within this Circuit] is read to be inconsistent with earlier case law, the earlier case law . . . controls.” (citation omitted)). Lending Corp., No. 12-0463, 2013 WL 12113414, at *3 (C.D. Cal. Feb. 11, 2013) (total asset freeze was appropriate where plaintiff showed revenue exceeding freezable assets and defendant did not meet its burden). There are multiple facts suggesting that at least half of Eonsmoke’s profits are attributable to a social media marketing strategy that was founded on infringement. To start, Tolmach testified that most, if not all, of Eonsmoke’s marketing strategy was based on social media. (Tolmach Tr. at 76:10–17.) Indeed, all of the social media posts produced in this case have used the Juul wordmark in some way.14 Additionally, Mr. Thomas’s report showed that Eonsmoke started the social media strategy in 2017, when revenues were $2.3 million, and that thereafter, revenues shot up to $30 million in 2018 and $90 million in 2019. (Thomas Rep. at 26.) While I cannot say that the social media marketing caused this jump, there is a highly suggestive correlation. To be sure, there is evidence of sales that did not derive from the infringing social media marketing. First, some sales were made by calling distributors directly. (Tolmach Tr. at 76:7–9). Second, some marketing was accomplished by having “influencers” promote the product, and there is no evidence that their posts including the infringing hashtags.15 (Id. at 76:6–11) Some Eonsmoke products were displayed in retail stores (Thomas Rep. at 37), so presumably some consumers might have bought them as a result of seeing them there. There is no evidence in the record of any other major forms of marketing, such as print advertising. Because neither side has yet quantified sales derived from direct marketing, influencer marketing, or on-the-spot sales, I cannot make a “dollar

14 The parties dispute whether Eonsmoke has engaged in spoliation, but there does not seem to be any dispute that, for whatever reason, Eonsmoke does not possess any further social media evidence. I may now have seen all social media posts that will be submitted in this case. 15 See generally “Influencer,” Merriam-Webster Dictionary (online ed. 2020), https://www.merriam-webster.com/dictionary/influencer (“a person who is able to generate interest in something (such as a consumer product) by posting about it on social media”). for dollar” estimate of the possible disgorgement award. Hoxworth, 903 F.3d at 199. Nevertheless, from these facts of record, I can infer that at least half of Eonsmoke’s revenue and profit can be tied to its central marketing strategy— infringing social media posts. Experience Hendrix, LLC v. Pitsicalis, No. 17 Civ. 1927 (PAE) (GWG), 2020 WL 3564485, at *7 (S.D.N.Y. July 1, 2020) (total revenue recoverable because it was reasonable to infer that all revenue was tied to infringement when use of the mark was the defendant’s primary business strategy), report & recommendation adopted, 2020 WL 4261818 (S.D.N.Y. July 24, 2020); see Covertech, 855 F.3d at 177 (noting that, in some cases, gross sales will all derive from infringing conduct). As a result, Juul’s requested asset freeze is equitably tailored because it likely covers less than what Juul could recover. That is enough to support an asset freeze representing half of conservative estimates of Eonsmoke’s profits. Necessity of the Asset Freeze I next determine whether, without the asset freeze, Juul will probably be unable to recover its likely disgorgement award. Juul I, 439 F. Supp. 3d at 358 (citation omitted). That is, I consider whether the assets will likely be consumed, dissipated, or fraudulently conveyed prior to final judgment. Elliott, 98 F.3d at 58. Previously, I found that Tolmach and Grishayev’s communications revealed “a clearly expressed intent to move or conceal assets, if and when a judgment is imminent or entered.” It is worth quoting the Court’s prior discussion and citation of the evidence on which that statement was based: On October 17, 2019, defendants produced a number of documents in discovery, including Skype messages between Mr. Tolmach and Mr. Grishayev sent in October and November 2018. In these messages, Mr. Tolmach and Mr. Grishayev express in no uncertain terms that they will never pay any judgment that may ultimately be entered in this action: Mr. Tolmach: Juul hitting us with a separate civil lawsuit for jury trial also. Mr. Tolmach: so one lawsuit for ITC for injunction and also one for civil lawsuit for penalties. even if we lose lawsuit, . . . Mr. Tolmach: even if we lose the lawsuit, I will never sign check to Juul or ever pay them Mr. Tolmach: will declare bankrupcy [sic] before that, so only attorneys getting paid here (DE 111-1 at 10). In a separate series of messages with a supplier in China, Mr. Grishayev voiced similar sentiments: Mr. Liu: worst case you stop this project and working on others Mr. Liu: even if you lose the lawsuit Mr. Grishayev: nah lawsuit i dont care Mr. Grishayev: just care for FDA Mr. Liu: you move your money away and apply for bankrupt [sic] Mr. Grishayev: yeah Mr. Liu: then start of a new company Mr. Grishayev: im not worried Mr. Grishayev: you can even still have same company lol Mr. Grishayev: even with bankrupt[cy] here in the usa (DE 111-1 at 15). Juul was understandably alarmed by these messages. Compounding that alarm were internal communications showing Mr. Tolmach and Mr. Grishayev communicating with each other to make a number of money transfers from their Eonsmoke business checking account—a Chase checking account with account number ending in 3070—to a Chase brokerage account, account number ending in 5906. This brokerage account is co-owned by Mr. Grishayev and Mr. Tolmach, and is nicknamed the “family office account” or the “family account.” (DE 111 at 14–15). The messages concerning these accounts suggest that defendants were—unusually—treating both the Chase checking account and the Chase brokerage account as Eonsmoke’s business accounts: Mr. Tolmach: gotta have like 5-10 mil on the books for Eonsmoke Mr. Tolmach: if you wanna be in mclane etc Mr. Tolmach: but then again checking account pays 0, it’s stupid to carry 800k in the checking Mr. Tolmach: guess we will just treat family office like the Eon checking acct for next few years Mr. Tolmach: and will move money back into eon checking topay huge taxes this time next year Mr. Grishayev: Exactly (DE 111-1 at 6 (emphasis added)). By letter filed January 23, 2020 (DE 154), Juul supplemented its presentation with another set of text messages, turned over in discovery, between Messrs. Grishayev and Tolmach: Mike T: Now juul is trying tactics like lobbying Chinese government and buying up all the nicotine Mike T: They’re mad creative Mike T: Moved another mil to fam office since we literally can’t send money away fast enough (Bates EON026958; DE 154-1) This exchange is in the same vein as the November 2018 messages already before the court. Counsel states that it dates from some six months later, in April 2019. (It is not clear whether the date April 25, visible on the printout, relates to this exchange or the following one.) Defendants represent that they are sweeping money out of the Chase checking account and into the brokerage “family” account to enjoy higher interest rates or investment returns. When they need funds to pay company taxes and expenses, however, they move money back from the brokerage account to the checking account. Juul replies that these texts will bear the interpretation that they are moving funds to preserve them from actions and threatened actions by Juul. Juul I, 439 F. Supp. at 347–48. At the time of Juul I, I did not find evidence of large movements or dissipation of funds. Id. at 361. I also found that Juul had not shown a likelihood of a large judgment, and that Eonsmoke still retained a large amount of assets, so an asset freeze was unnecessary. Id. Now, several new facts, taken together, show that an asset freeze is warranted. First, Tolmach and Grishayev have consumed a fair amount of assets since Juul I. Tolmach spent nearly $3 million on a house, renovations, and furniture. (DE 223-1; 227-6; 231-2, Ex. 11, 12.) There have also been unexplained transfers and small-scale consumption. (DE 227-1; 227-2.) Because “consumption” of assets may warrant an asset freeze, Elliott, 98 F.3d at 58, this significant amount of personal spending supports an asset freeze. Additionally, these purchases all qualify as personal or even luxury consumption—further indicating that defendants may be spending money for the sake of spending money prior to judgment.16 Second, Eonsmoke moved around 90% of assets from the family office account to new, “substitute accounts” held by Tolmach and Grishayev individually. (DE 167-2, at 16–17; 181; 223.) Although Eonsmoke explained that the new accounts provided “more advantageous account management and investment income” (DE 167, at 1), when pressed by Juul for an admission

16 As I noted at oral argument, some of the purchases, like the Los Angeles property and the furniture, might be recouped by Juul. However, liquidating Eonsmoke’s non-cash assets is not a simple task and might or might not result in recovery of the full purchase price. Indeed, perhaps for this reason, a court in this District held that a defendant’s down payments on properties supported an asset freeze. Wilmington Savings Fund Soc’y v. Otieno-Ngoje, Civ. No. 16-5631, 2016 WL 6821086, at *2 (D.N.J. Nov. 17, 2016). Moreover, as the Supreme Court recently explained, when a plaintiff has a right to disgorgement, the infringer can be said to hold the plaintiff’s profits in trust. Liu, 140 S. Ct. at 1944–45. Accordingly, the assets are not really Eonsmoke’s to spend for personal reasons, and the equities support freezing them to prevent unnecessary consumption. that these accounts would serve as business accounts, Eonsmoke would not say so. (DE 232-2). While Eonsmoke’s prior transfers could have been explained away more easily, additional transfers create more suspicion. Considering that the purpose of a preliminary injunction is to maintain the status quo, Acierno v. New Castle County, 40 F.3d 645, 647 (3d Cir. 1994), the new accounts and transfers weigh in favor freezing the assets so that no more movements can occur. Compare Chi. Title Ins. Co. v. Lexington & Concord Search & Abstract, LLC, 513 F. Supp. 2d 304, 320 (E.D. Pa. 2007) (no asset freeze when alleged conveyance of assets only represented a small portion of his net assets). Third, Eonsmoke is no longer in business. (DE 231-1; 182-1 ¶ 6; Tolmach Tr. at 305:25–06:4.) As a result, Eonsmoke likely has no legitimate business reason to spend or move any of its assets. Eonsmoke has little or no incentive to maintain business-related accounts, beyond the need to wind up any pending tax or other obligations. There is a real risk that it would transfer its assets to individuals or spend those assets. See Gerardi v. Pelullo, 16 F.3d 1363, 1373 (3d Cir. 1994) (corporation’s “insubstantial existence” supported asset freeze). Indeed, as explained, Eonsmoke has already done so, giving reason to think it will do so again unless restrained. Fourth and finally, Eonsmoke’s diminution of assets seems to be outpacing its quarterly reporting obligations. Eonsmoke’s last quarterly report to the Court, in September 2020, showed some $25 million in assets. (DE 227.) In its brief filed in November, Eonsmoke represented that it possessed around $20 million. (Eonsmoke Opp. at 4.) Eonsmoke’s next quarterly report is not yet due, and it may provide an explanation for the difference. (For example, earlier reports from Eonsmoke noted it was paying a significant amount of taxes. (DE 167)). Nevertheless, there does seem to have been a $5 million drop in assets over a short period. Accordingly, an asset freeze is needed to prevent any further changes to the status quo and ensure that assets are used for legitimate purposes. See Acierno, 40 F.3d at 653; Elliott, 98 F.3d at 61. Given Tolmach’s and Grishayev’s clearly expressed intent, combined with additional consumption or movement of assets, and Eonsmoke’s defunct status, an asset freeze is warranted. Juul has made an adequate showing regarding the size of a potential disgorgement award, and I am satisfied that Juul would be irreparably harmed absent an asset freeze because it would lose part of its equitable remedy. D. Remaining Injunction Factors I consider the remaining preliminary injunction factors, supplementing my findings from Juul I. Both favor an asset freeze. First, I must identify the possible harm Eonsmoke would face with an asset freeze and balance that harm against Juul’s potential injury without an asset freeze. Ramsay, 968 F.3d at 263. Of course, assuming the assets rightfully belong to Juul, there is no legally cognizable harm at all, but there has been no final judgment to that effect. Now “encumbering at least tens of millions of dollars of [Eonsmoke’s] assets” would clearly harm defendants. Hoxworth, 903 F.3d at 207. But two facts cut against a finding of significant harm: (1) Eonsmoke is no longer in business, so its assets are not needed for any legitimate expenses, and an asset freeze would not imperil its business; and (2) Eonsmoke’s principals have spent the assets on luxury and personal expenses, so there is little injury in having to forgo such discretionary spending, Elliott, 98 F.3d at 59 (“The inability to spend . . . on primarily luxury expenses does not strike us as a hardship.”). On the other side of the balance, Juul has made strong showings that (1) it will succeed on its wordmark claim; (2) that claim is worth a substantial amount of money, quite possibly more than Eonsmoke’s current assets; and (3) its prospective judgment is legitimately at risk given Eonsmoke’s conduct. Accordingly, the hardship to defendants from these proposed restrictions is outweighed by the harm Juul faces if Eonsmoke cannot pay a judgment. Id. Next, I assess whether the public interest supports issuance of an injunction. Ramsay, 968 F.3d at 263. “[P]ublic interest in a trademark case is most often a synonym for the right of the public not to be deceived or confused.” Kos Pharms., Inc. v. Andrx Corp., 369 F.3d 700, 730 (3d Cir. 2004). Here, the infringing conduct has already stopped, so there is no further confusion to prevent. Still, “the public interest is advanced by recognition of property interests in trademarks.” Bill Blass Ltd. v. SAZ Corp., 751 F.2d 152, 156 (3d Cir. 1984). And that interest extends to disgorgement of profits from infringement. An asset freeze would ensure that Eonsmoke can pay a likely disgorgement award, which, in turn, effectively recognizes that Eonsmoke’s profits were never its rightful property. Accordingly, ensuring that Eonsmoke cannot avoid a judgment that would recognize Juul’s “property interests in trademarks” serves the public interest. See id. Put another way, allowing Eonsmoke to make itself judgment-proof would undermine the public interests protected by trademark law. See Ramsay, 968 F.3d at 263 (public interest is served by requiring entities to comply with the law); PNY Techs. Inc. v. Salhi, Civ. No. 12-4916, 2016 WL 4267940, at *3 (D.N.J. Aug. 10, 2016) (“It is not in the public interest to allow defendants to make themselves judgment-proof.”). Thus, the remaining preliminary injunction factors also favor an asset freeze. E. Scope of Injunction Having concluded that an asset freeze is warranted, I consider its appropriate scope. An asset freeze should not impair any more assets than necessary to maintain the status quo and ensure that Eonsmoke can pay the likely disgorgement award. See Elliott, 98 F.3d at 60. More generally, injunctions should be no more burdensome than necessary, and a district court should consider principles of fairness to the enjoined party. SEC v. Gentile, 939 F.3d 549, 559–60 (3d Cir. 2019) (citations omitted), cert. denied, 140 S. Ct. 2669 (2020). At bottom, “the district court has very broad power to fashion a remedy appropriate to deal with the factual situation before the court.” Lewis v. Kugler, 446 F.2d 1343, 1352 (3d Cir. 1971). It appears from the record and the quarterly reporting that Eonsmoke’s assets are now spread among several accounts: the original checking account and “family office” brokerage account (which were the subject of Juul I), as well as the substitute individual accounts. Accordingly, Eonsmoke will be enjoined from moving any funds out of those accounts, which are specifically listed in the accompanying order, as well as traceable funds in any accounts to which they have been transferred. It would be punitive, and therefore improper, to use this asset freeze to strip Tolmach and Grishayev of the ability to support themselves, based on no more than a preliminary finding of likely liability. See Gentile, 939 F.3d at 559– 60. So, while the accounts generally will be frozen, Tolmach and Grishayev will each be permitted to withdraw up to a total of $500,000 from the accounts, upon letter application for the Court’s authorization, which will not unreasonably be withheld. Additionally, Eonsmoke may move the Court, for the duration of the asset freeze, for permission to make a transfer from the accounts, and the Court will determine whether there is a legitimate business or personal reason for the transfer. To ensure compliance, my previous order requiring quarterly reporting will continue in effect. Finally, Juul asks that I also issue injunctive relief against the financial institutions that maintain the accounts. (Juul Brf. at 19.) Those institutions are not parties to this case, but under the All Writs Act, 28 U.S.C. § 1651(a), I have some power to enjoin them to ensure compliance with the asset freeze. United States v. N.Y. Tel. Co., 434 U.S. 159, 172–44 (1977); see also Berger v. Zeghibe, 666 F. App’x 119, 123 (3d Cir. 2016) (relying on New York Telephone to affirm an asset freeze as applied to a non-party). More generally, of course, an injunction may reasonably bind those who have notice of it. I will therefore direct that a copy of the accompanying court order be served on the financial institutions involved. F. Bond Finally, a prerequisite to issuance of injunctive relief is that Juul post a bond “in an amount that the court considers proper to pay the costs and damages sustained by any party found to have been wrongfully enjoined or restrained.” Fed. R. Civ. P. 65(c); see also Hope v. Warden York Cnty. Prison, 972 F.3d 310, 322 (3d Cir. 2020). Juul asks me to waive the bond requirement because, in its view, the equities overwhelmingly favor itself. (Juul Brf. at 5.) The Third Circuit holds, however, that instances where a district court may waive the bond requirement are rare, and the court has “never excused a District Court from requiring a bond where an injunction prevents commercial, money-making activities.” Zambelli v. Fireworks Mfg. Co. v. Wood, 592 F.3d 412, 326 (3d Cir. 2010). The exception which Juul invokes does not apply here, see id., and moreover, Juul has not explained how posting a bond would be a hardship. In setting the amount of bond, I will assume that two years is a reasonable amount of time for completion of this litigation.17 Because Juul is a financially substantial entity, because I have authorized reasonable withdrawals upon application to the Court, and because the amount of the bond can be revisited at any time should any new basis for insecurity arise, that should be sufficient. For its part, Eonsmoke asks that the bond be set at $50 million, the total disgorgement which Juul seeks. (Eonsmoke Opp. at 5.) But a bond is only justified to the extent necessary to secure the price of a wrongfully issued

17 One way the preliminary injunction might be found to have been unwarranted is by way of a final judgment in this action. The other, of course, would be a successful appeal. This asset freeze, as a preliminary injunction, is appealable on an interlocutory basis. See 28 U.S.C § 1292; Def. Distrib. v. Att’y Gen. of N.J., 972 F.3d 193, 198 (3d Cir. 2020). Should that occur, the Court of Appeals’ mandate would likely issue within one year. See The Bar Association for the Third Federal Circuit, U.S. Court of Appeals for the Third Circuit Practice Guide at 25 (2012), http://thirdcircuitbar.org/documents/third-circuit-bar-practice-guide.pdf (estimating the average civil appeal to take 10.4 months from notice of appeal to disposition). injunction. Sprint Commc’ns Co. v. CAT Commc’ns Int’l, Inc., 335 F.3d 235, 240 (3d Cir. 2003). By Eonsmoke’s own account, there is only about $20 million in assets to freeze. That, not $50 million, is the starting point for analysis of the bond amount. Moreover, the $20 million, if frozen, will not disappear or be lost. Even if an asset freeze were ultimately found wrongful, Eonsmoke would have been temporarily deprived of the use of $20 million, not of the $20 million itself. Eonsmoke itself is no longer in operation and has identified no pressing need to fund ongoing operations. Should there be any outstanding tax or other corporate obligations, it may apply for release of funds. The parties directly affected by any deprivation would presumably be Mr. Tolmach and Mr. Grishayev. By their own admission, after expressing an intent to avoid paying any judgment, these two moved Eonsmoke’s funds to a joint account which they called the “office account.” They later split those funds for deposit to two individual, interest-bearing accounts. As noted above, I have already authorized an exemption from the freeze totaling $1 million, in recognition of the preliminary nature of the plaintiff’s showing and these two individuals’ need to meet personal expenses. That exemption negates one claimed source of loss, i.e., the deprivation of funds necessary for the activities of life. It represents cash—from defendants’ point of view, a form of security far superior to a bond. The time value of the money might constitute a potential loss from a wrongfully granted injunction. That loss is offset, however, by the fact that the funds will presumably remain precisely where the defendants have placed them, in the individual accounts or elsewhere. If, for example, the funds were accruing interest, they will continue to do so. Taking all of the above into account, I will require a bond of $1 million, representing approximately 5% of the frozen funds. III. CONCLUSION For the reasons set forth above, Juul has made a sufficient showing for a preliminary injunction in the form of an asset freeze. Specifically, it has shown a likelihood of success on its wordmark claim, that the profits derived from wordmark infringement far surpass the remaining assets held by Eonsmoke, and that there is a sufficient risk that Eonsmoke will deplete those assets before final judgment. As a result, Eonsmoke will be prohibited from transferring any funds from its accounts, subject to certain exceptions explained above and included in the accompanying order. A separate order will issue. Dated: December 22, 2020 /s/ Kevin McNulty ___________________________________ Hon. Kevin McNulty United States District Judge