govinfo:USCOURTS-cand-3_22-cv-07197-3

1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
 
 
 
 
 
 
UNITED STATES DISTRICT COURT 
NORTHERN DISTRICT OF CALIFORNIA 
 
LA CANADA VENTURES, INC., 
Plaintiff, 
v. 
 
MDALGORITHMS, INC., 
Defendant. 
 

 
 
ORDER GRANTING IN PART AND 
DENYING IN PART THE PARTIES’ 
POST-TRIAL MOTIONS 
 
 
Plaintiff La Canada Ventures is a health and beauty company that markets its products 
under a variety of trademarks that begin with “MD.” La Canada sued MDalgorithms, another 
health and beauty company, averring that MDalgorithms infringed on its trademarks by offering 
its signature hair regrowth solution under the name “MDhair.” La Canada tried its claims to a jury, 
and it won. The jury found that MDalgorithms infringed one or more of La Canada’s registered 
trademarks and one or more of La Canada’s unregistered trademarks. It awarded La Canada 
$1,900,000 in actual damages. 
The parties now move for various forms of post-trial relief. La Canada asks for a 
permanent injunction, enhanced damages, disgorgement of profits, judgement on partial findings 
on two of MDalgorithms’ counterclaims, and attorney’s fees. MDalgorithms has renewed its 
motion for judgment as a matter of law on the trademark infringement claims and has moved for a 
new trial. 
Though a jury verdict is never lightly set aside, this one must be—at least in large part. 
There is a reasonable evidentiary basis for the jury’s liability finding, but there is not one for the 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

2 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
bulk of its damages award. The evidence supports a maximum damages award of $33,205, not 
$1.9 million. Therefore, MDalgorithms’ motion for judgment as a matter of law is granted in part. 
Its motion for a new trial as to liability is denied, and its motion for a new trial as to damages is 
conditionally granted for the same reasons its motion for judgment as a matter of law as to 
damages is granted. See Fed. R. Civ. P. 50(c)(1). All of La Canada’s motions are denied, save one. 
Its motion for judgment on partial findings as to MDalgorithms’ eighth and ninth counterclaims—
both for unfair competition—is granted because MDalgorithms conceded that it suffered no 
competitive injury. 
I. BACKGROUND 
This is a trademark infringement case contested between two purveyors of health and 
beauty products. The first is La Canada Ventures, Inc. La Canada, which was founded by Dr. 
Susan Lin, began offering health and beauty products under marks beginning with “MD” in 2006. 
Several of those marks were registered with the USPTO by Dr. Lin and subsequently assigned to 
La Canada. La Canada has also offered products using various unregistered trademarks since as 
early as 2007. 
The second is MDalgorithms. Founded roughly a decade ago, MDalgorithms uses 
algorithmic techniques to customize health and wellness solutions for its customers. Beginning in 
2021, MDalgorithms began to offer a hair restoration solution under the name MDhair. Between 
January 2022 and August 2025, MDalgorithms distributed more than 290,000 shipments of 
MDhair to consumers within the United States. 
This dispute began in April 2022, when a customer contacted La Canada about MDhair. La 
Canada sent a cease-and-desist letter to MDalgorithms, demanding that it stop using the MDhair 
mark because it was causing confusion in the marketplace. MDalgorithms refused, so La Canada 
filed this lawsuit. La Canada claimed that MDalgorithms’ decision to brand its haircare solution 
MDhair violated La Canada’s trademark rights under the Lanham Act and analogous California 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

3 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
unfair competition law.1 See 15 U.S.C. § 1114; Cleary v. News Corp., 30 F.3d 1255, 1262–63 (9th 
Cir. 1994) (explaining that “actions pursuant to California Business and Professions Code § 
17200 are ‘substantially congruent’ to claims made under the Lanham Act” (quoting Academy of 
Motion Picture Arts & Sciences v. Creative House Promotions, Inc., 944 F.2d 1446, 1457 (9th Cir. 
1991))). La Canada invoked the protection of seven registered trademarks: MD, MD 101, MD 
Lash Factor, MD Intimate Restore, MD by Susan F. Lin, M.D., MD Wellness by Susan F. Lin 
M.D., and MD Factor. It also invoked the protection of two unregistered trademarks: MD NUTRI 
HAIR and MD HAIR. 
To succeed in its trademark infringement claims, La Canada had to show three basic 
things: (1) that the asserted marks are valid and protectible, (2) that it owned those trademarks, and 
(3) that MDalgorithms used marks similar or identical to La Canada’s marks in a manner likely to 
cause confusion among consumers. See Lerner & Rowe PC v. Brown Engstrand & Shely LLC, 119 
F.4th 711, 718 (9th Cir. 2024). Because La Canada claimed ownership to registered and 
unregistered trademarks, it sought to prove validity in two ways. As to the former marks, it 
asserted that validity and protectability were established by the registration itself. As to the latter 
marks, La Canada argued that the marks are “suggestive.” That is, because they start with “MD,” 
they imply a characteristic or quality about the product in the minds of consumers: namely, that La 
Canada’s products are medical-grade solutions created by a physician. Suggestive marks are 
“inherently distinctive,” making them valid and protectible. Two Pesos, Inc. v. Taco Cabana, Inc., 
505 U.S. 763, 768 (1992). In the alternative, La Canada posited that its marks had acquired a 
“secondary meaning” among consumers in the health and beauty space. Said otherwise, even if the 
marks were generic in a vacuum, customers came to associate marks beginning with “MD” with 
La Canada through years of consistent use. 
To show ownership, La Canada had to demonstrate that it began to use the marks in 
 
1 La Canada also brought trademark infringement claims related to MDalgorithms’ skincare 
solution, MDacne. Summary judgment was granted as to those claims in favor of MDalgorithms. 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

4 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
commerce before MDalgorithms began to use its allegedly infringing marks. It offered the 
testimony of Dr. Lin, who explained that La Canada developed the product offered under one of 
the registered marks, MD Lash Factor, as early as 2007. She went on to testify that by 2013 at the 
latest—still several years before MDalgorithms was founded—La Canada was selling a shampoo 
and conditioning product using the mark MD HAIR. 
The final element has two component parts. First, the plaintiff must show that the 
defendant used marks similar or identical to its marks. That was not a tall order here. As 
explained, MDalgorithms offered a hair restoration product under the mark MDhair. That differs 
from La Canada’s MD HAIR mark only in that the former does not contain a space and uses lower 
case letters for the word “hair.”2 For good measure, La Canada’s MD NUTRI HAIR mark differs 
from MDhair as does MD HAIR plus it contains one extra word: NUTRI. 
The second component requires the plaintiff to show that the alleged infringement was 
likely to cause confusion in the marketplace. The nature of the confusion turns on the plaintiff’s 
theory of the case. Here, La Canada advanced a theory of forward confusion. In a paradigmatic 
forward confusion case, “the well-known mark goes after a look-alike, sound-alike, feel-alike 
unknown which is trying to cash in on the famous mark’s goodwill.” Dreamwerks Prod. Grp., Inc. 
v. SKG Studio, 142 F.3d 1127, 1128 (9th Cir. 1998). Imagine that an upstart retailer starts to sell 
luxury goods under the mark Hermène, complete with orange-accented advertising centered 
around equestrian imagery. Hermès—the well-known French luxury brand that advertises its 
goods in much the same way—could state a strong forward confusion claim because consumers of 
luxury goods might reasonably look at Hermène’s products and conclude they come from, are 
sponsored by, or are in some way affiliated with Hermès. 
That means, here, La Canada bore the burden of demonstrating that MDalgorithms’ use of 
 
2 It is somewhat unclear from the record whether La Canada’s mark is “MD HAIR” or “MD Hair.” 
La Canada’s papers use the capitalized version, but its public-facing website uses the lower-case 
version. Regardless, the jury was permitted to conclude that MDalgorithms’ “MDhair” mark was 
similar to either. 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

5 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
the MDhair mark was likely to confuse consumers into thinking that La Canada was the source of 
MDhair. The Ninth Circuit has provided a list of eight factors to be assessed in determining the 
likelihood of confusion between related goods, see AMF Inc. v. Sleekcraft Boats, 599 F.2d 341, 
348–49 (9th Cir. 1979), but the relative contribution of each factor to the analysis will depend on 
the facts of each case, see Marketquest Grp., Inc. v. BIC Corp., 862 F.3d 927, 934 (9th Cir. 2017) 
(“The Sleekcraft ‘analysis is pliant, illustrative rather than exhaustive, and best understood as 
simply providing helpful guideposts.’” (quoting Fortune Dynamic, Inc. v. Victoria’s Secret Stores 
Brand Mgmt., Inc., 618 F.3d 1025, 1030 (9th Cir. 2010))). They are: 
(1) strength of the mark; 
(2) proximity of the goods; 
(3) similarity of the marks; 
(4) evidence of actual confusion; 
(5) market channels used; 
(6) type of goods and the degree of care likely to be exercised by the purchaser; 
(7) defendant’s intent in selecting the mark; and 
(8) likelihood of expansion of the product lines. 
Sleekcraft, 599 F.2d at 348–49 (hereafter, the “Sleekcraft factors”). La Canada packaged those 
factors into two arguments. First, it claimed that consumers were likely to be confused as to the 
origin of MDhair because La Canada and MDalgorithms used exceedingly similar marks, targeted 
the same customers, and both sold their products over the internet. 
 Second, it claimed to have evidence of significant actual confusion. To be clear, a 
trademark infringement plaintiff need not prove that confusion actually occurred. Rather, evidence 
of actual confusion is probative of the fact that the defendant’s use of the mark was of a kind likely 
to create confusion among consumers, which is all that is legally relevant. La Canada presented its 
evidence of actual confusion through its sole employee, Cassandra Havea. Havea has been 
employed at La Canada since 2007, and one of her responsibilities is to monitor communications 
from customers. She testified that, in April 2022, she received an email from someone about 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

6 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
MDhair—the MDalgorithms product. These emails continued until two days before she took the 
stand. 
 Some of these errant inquiries were manifestly made by MDalgorithms customers. For 
instance, one customer sent Havea proof of payment on MDhair.co, which is the domain through 
which MDalgorithms sells MDhair. Others took some inference. In one case, a customer called 
Havea but was unable to provide an order number. She then asked the customer what the color of 
the product was and he answered “blue bottle.” That was enough, in her telling, to deduce that 
inquiry was coming from an MDalgorithms customer because MDalgorithms offers MDhair in a 
blue bottle. 
 In still other cases, there was no evidence at all that the inquiry was from an MDalgorithms 
customer. Havea testified that if she received a communication from someone who was not known 
as a La Canada customer, she assumed that person was an MDalgorithms customer that was 
confused about the source of MDhair. For example, one customer contacted La Canada and simply 
said he had not received his parcel. Havea admitted that the message did not mention MDhair or 
MDalgorithms, but she nonetheless concluded that he must have been an MDalgorithms customer 
because “when you put [in] the customer name, they never order[ed] from La Canada.” Tr. 288:2–
4. In another marked as an instance of actual confusion, the customer simply said “hi.” Tr. 285:1–
9. Even more confusingly, Havea assumed that some individuals that contacted La Canada were 
MDalgorithms customers even if there was evidence that they had previously been La Canada 
customers. In one case, Havea testified that she marked an individual that complained about not 
getting a package as an MDalgorithms customer even though he had “been a La Canada customer 
for over two years.” Tr. 292:13–20. In all, Havea logged 377 instances of actual confusion, but in 
only 73 of those instances was there some indication that the communication was coming from an 
MDalgorithms customer. 
 MDalgorithms moved for judgement as a matter of law after the close of La Canada’s 
case-in-chief. See Fed. R. Civ. P. 50(a). MDalgorithms’ headline argument was that no reasonable 
juror could find for La Canada on its trademark infringement claims because the Sleekcraft factors 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

7 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
weighed decisively against La Canada. MDalgorithms also argued that equity barred recovery 
because La Canada facilitated any confusion by employing search engine optimization and domain 
redirection techniques that sent consumers looking for MDalgorithms to La Canada’s website. 
MDalgorithms’ motion was denied subject to renewal after the verdict, but it was close. See Tr. 
536:24–537:7 (“Court: [I]t’s pretty thin for the plaintiffs . . . I don’t think there’s much there that 
shows [there was confusion about the source of MDhair].”). 
 The jury found MDalgorithms liable for infringing at least one of La Canada’s registered 
trademarks and liable for infringing at least one of La Canada’s unregistered trademarks. It 
awarded La Canada $760,000 for the former and $1,140,000 for the latter. MDalgorithms timely 
renewed its motion for judgment as a matter of law, see Fed. R. Civ. P. 50(b), and moved for a 
new trial on the same basis, see Fed. R. Civ. P. 59(a)(1)(A). Buoyed by its success before the jury, 
La Canada filed five post-trial motions of its own: for (1) a permanent injunction, (2) 
disgorgement of profits, (3) enhanced damages, (4) judgment on partial findings on two of 
MDalgorithms’ counterclaims, (5) and attorney’s fees. 
II. DISCUSSION 
a. MDalgorithms’ Motions 
Under Rule 50(a), a party may move for judgment as a matter of law on any issue on 
which the opposing party has been fully heard. See Fed. R. Civ. P. 50(a). If that motion is denied, 
the moving party may renew it at the close of evidence. See Fed. R. Civ. P. 50(b). The inquiry is 
identical to that applied in a motion for summary judgment under Rule 56. See Reeves v. 
Sanderson Plumbing Prods., Inc., 530 U.S. 133, 150 (2000). Namely, a motion for judgment as a 
matter of law may only be granted “if, under the governing law, there can be but one reasonable 
conclusion as to the verdict.” Winarto v. Toshiba Am. Elecs. Components, Inc., 274 F.3d 1276, 
1283 (9th Cir. 2001). In making that assessment, all the evidence must be considered, all 
inferences must be drawn in favor of the nonmoving party, and no credibility determinations are 
permitted. See Reeves, 530 U.S. at 150. However, as at summary judgment, the nonmoving party 
cannot defeat a motion for judgment as a matter of law by pointing to a mere scintilla of evidence 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

8 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
that could support a verdict in its favor. See In re Oracle Corp. Sec. Litig., 627 F.3d 376, 387 (9th 
Cir. 2010). 
Under Rule 59, a new trial may be granted “for any reason for which a new trial has 
heretofore been granted in an action at law in federal court.” Fed. R. Civ. P. 59(a)(1)(A). Rule 59 
does not specify the grounds upon which a new trial will be granted; rather, the district court is 
“bound by those grounds that have been historically recognized.” Zhang v. Am. Gem Seafoods, 
Inc., 339 F.3d 1020, 1035 (9th Cir. 2003). “Historically recognized grounds include, but are not 
limited to, claims ‘that the verdict is against the weight of the evidence, that the damages are 
excessive, or that, for other reasons, the trial was not fair to the party moving.’” Molski v. M.J. 
Cable, Inc., 481 F.3d 724, 729 (9th Cir. 2007) (quoting Montgomery Ward & Co. v. Duncan, 311 
U.S. 243, 251 (1940)). The standard for granting a new trial is lower than that for granting 
judgment as a matter of law. See Experience Hendrix L.L.C. v. Hendrixlicensing.com Ltd, 762 
F.3d 829, 841 (9th Cir. 2014). However, because MDalgorithms advances the same arguments in 
support of both motions, they are assessed in parallel and addressed separately only where 
necessary. 
i. The Jury’s Liability Verdict 
A. Likelihood of Confusion 
MDalgorithms’ first argument is that no reasonable juror could have concluded that 
MDalgorithms’ use of the MDhair mark was likely to confuse consumers into thinking MDhair 
came from La Canada. As explained, the likelihood of confusion between two products is assessed 
using the Sleekcraft factors. See, 599 F.2d at 348–49. La Canada leaned most heavily into the 
fourth factor: evidence of actual confusion. That evidence was presented exclusively through 
Cassandra Havea, who testified that she cataloged 377 instances of actual confusion between April 
2022 and March 2026. However, only 73 of those logged instances came from a verified 
MDalgorithms customer, and only 41 came from someone within the United States. 
 MDalgorithms argues that no reasonable juror could have concluded that this was anything 
more than evidence of de minimis confusion. The evidence revealed that, between late 2021 and 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

9 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
June 2025 (which is not exactly the same period as that covered by Havea’s log, though it is 
close), MDalgorithms shipped 290,000 units of MDhair to consumers within the United States. 
That produces a confusion rate of 0.13% at the high end (using a total of 377 instances of actual 
confusion) and 0.014% at the low end (using a total of 41 instances of actual confusion). 
 In that way, MDalgorithms argues that this case is similar to Lerner & Rowe PC. See 119 
F.4th at 719–21. Lerner & Rowe PC involved a dispute between two personal injury law firms, 
one of which accused the other of infringing on its trademark through “conquesting.” See id. at 
717. Conquesting is when a lesser-known firm purchases its better-known competitor’s branded 
keywords so that customers who search for the better-known firm online see results for the lesser-
known firm. See id. To show that the conquesting had confused customers, the plaintiff in Lerner 
& Rowe PC presented evidence of 236 phone calls that the defendant received in which the caller 
identified the plaintiff by name. See id. at 719. Google data revealed that, during the relevant 
period, 109,322 searches for the plaintiff firm showed an advertisement for the defendant firm. See 
id. at 720. “Evidence of 236 instances of actual confusion, therefore, constitute[d] only 0.216% of 
the total number of users exposed to the challenged advertisements.” Id. 
The district court dismissed that evidence of actual confusion as de minimis at summary 
judgment, and the Ninth Circuit affirmed. See 119 F.4th at 720. Much of the panel’s analysis 
focused on how large and how ascertainable the denominator—that is, the total number of 
opportunities for confusion—was. It explained that in cases in which the denominator is uncertain, 
“we see the tip of an iceberg and have no ability to speculate about how much lies below the 
surface.” Id. By contrast, when the denominator is known, “no speculation is necessary—we can 
see the entire iceberg.” Id. That was the case in Lerner & Rowe PC, where the Google data 
allowed the factfinder to “discern with a high degree of precision the proportion of all consumers 
who were actually confused.” Id. 
Discerning that precise quotient is much more difficult here. As La Canada points out, the 
appropriate denominator is only those consumers who encountered both trademarks in the 
marketplace. Those who never came across La Canada ought not be counted in the calculation 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

10 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
because they never had an opportunity to be confused. MDalgorithms takes this as some kind of 
confession: that if only a fraction of MDalgorithms’ customers encountered La Canada, there can 
be no likelihood of forward confusion. That is wrong. What matters is whether those who do see 
both marks are likely to be confused about the source of the junior user’s products; there is no 
requirement that there is perfect overlap between the two company’s consumers or that all 
consumers of the junior user know about the senior user. That is because the actual confusion 
factor is merely an empirical gauge used to answer the pertinent legal question: whether the junior 
user used the mark in a manner likely to cause confusion among consumers. 
Second, the documented instances of actual confusion may underestimate the likelihood of 
confusion here because they capture only those consumers who were both confused and had some 
question about or dissatisfaction with MDhair. Said otherwise, a customer who thought that 
MDhair was a La Canada product but who had no reason to complain would not show up in the 
data, even though that person was confused. (How that would bear on damages is a separate 
question.) To contextualize the instances of actual confusion properly here, the jury would have 
had to know how many MDhair consumers encountered both MDalgorithms and La Canada and 
wanted to contact the source provider. 
In the absence of this context, the jury was permitted to credit the evidence of actual 
confusion in finding that consumers were likely to be confused about the source of MDhair. The 
Ninth Circuit allowed as much in Ironhawk Technologies, Inc. v. Dropbox, Inc. 2 F.4th 1150, 
1165–66 (9th Cir. 2021). There, Ironhawk, which sold a software branded “SmartSync” sued 
Dropbox, which offered a software suite with a feature called “Smart Sync,” on a theory of reverse 
confusion. See id. at 1159–60. To demonstrate a likelihood of confusion, Ironhawk relied in part 
on evidence of actual confusion. See id. at 1165–66. That evidence consisted of the testimony of 
two individuals: the Ironhawk CEO and a third-party sales representative. See id. Both individuals 
testified that Ironhawk’s customers and potential customers were confused—some thought that 
they already bought Ironhawk’s “SmartSync” when they had actually bought the suite that 
included Dropbox’s “Smart Sync”; others asked about the relationship between Ironhawk’s 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

11 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
SmartSync and Dropbox. See id. Though the Ninth Circuit expressed some skepticism that the 
jury would credit this evidence at trial, it permitted it to do so. See id. Without any evidence as to 
the denominator that could render the actual confusion evidence de minimis, the instances of 
actual confusion were “evidence a reasonable jury could rely on to support a finding of actual 
confusion or when assessing a likelihood of confusion under the totality of the circumstances.” Id. 
at 1166; see Lerner & Rowe PC, 119 F.4th at 721 (distinguishing Ironhawk on the basis that, 
there, the panel “weighed individual instances of confusion without the benefit of knowing the 
total number of opportunities consumers had for confusion”). The jury here exercised the same 
prerogative. 
Nor do the other Sleekcraft factors weigh decisively enough in MDalgorithms’ favor to 
require the liability finding to be set aside. The jury was entitled to conclude that the first factor, 
strength of the mark, supported La Canada. Dr. Lin testified that La Canada has been using a 
family of MD-formative marks for almost two decades and has spent hundreds of thousands of 
dollars on advertising, imbuing the marks with substantial commercial strength—or at least the 
jury could so conclude. MDalgorithms’ motion rightly focuses on the countervailing evidence it 
presented at trial, most notably its recitation of the myriad other marks that use “MD” and the 
natural association of “MD” with “medical doctor” as opposed to La Canada. Many a jury would 
have found that evidence demonstrated the marks’ lack of conceptual strength. Either this one did 
not or—much more likely—it decided that it did not outweigh the other factors that pointed in La 
Canada’s direction. That conclusion was reasonable. 
The jury could have reasonably concluded that several of the other factors weighed in La 
Canada’s favor too. Factor three asks how similar the marks are. As explained, MDalgorithms’ 
MDhair mark differs from the closest La Canada mark, MD HAIR, only in capitalization and the 
addition of a space. It further differs from La Canada’s MD NUTRI HAIR mark through 
capitalization, the addition of a space, and an extra word. Factor two asks how similar the goods 
are. Here, both support hair growth and health. The jury’s apparent decision to credit those factors 
in evaluating the totality of the circumstances was not unreasonable. 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

12 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
B. Willful Infringement 
MDalgorithms next argues that there was insufficient evidentiary support for the jury’s 
conclusion that its infringement was willful. Two preliminary points must be made before 
assessing the merits of this argument. First, a finding of willfulness merely bears on the remedies 
available to an infringement plaintiff—it is not an element of liability nor a requirement to recover 
actual damages. See Romag Fasteners, Inc. v. Fossil, Inc., 590 U.S. 212, 219 (2020) (explaining 
that “a trademark defendant’s mental state is a highly important consideration in determining 
whether an award of profits is appropriate”); Love v. Associated Newspapers, Ltd., 611 F.3d 601, 
615 (9th Cir. 2010) (discussing the relationship between willful infringement and an award of 
attorney’s fees). Thus, while willfulness is discussed in the context of MDalgorithms’ Rule 50(b) 
and 59 motions because that is where MDalgorithms raised it, MDalgorithms’ argument is best 
understood as a challenge to La Canada’s request for disgorgement and attorney’s fees. The 
adjudication of those requests thus incorporates the analysis in this section. 
Second, La Canada argues vigorously that this argument—as well as MDalgorithms’ 
arguments regarding damages and the alleged reverse confusion theory—are procedurally barred 
because MDalgorithms did not raise either in its original motion for judgment as a matter of law 
under Rule 50(a). Even if that is true, the merits of those arguments must still be addressed 
because MDalgorithms raised them all in its Rule 59 motion, and it was under no obligation to 
make those arguments in its Rule 50(a) motion to preserve them for that purpose. Moreover, for 
reasons explained below, these arguments do not compel granting either motion, so La Canada 
suffers no prejudice from considering them in the Rule 50(b) context. 
The jury was instructed that willfulness required a showing either that MDalgorithms 
intended to deceive consumers as to the origins of MDhair or that it was willfully blind to the 
nature of its conduct. See Atari Interactive, Inc. v. Redbubble, Inc., 546 F. Supp. 3d 883, 886 
(N.D. Cal. 2021), aff’d in part, appeal dismissed in part, No. 21-17062, 2023 WL 4704891 (9th 
Cir. July 24, 2023). MDalgorithms’ motion focuses exclusively on the first of the two paths to 
show willfulness, and it largely retreads the grounds on which its objection to the jury’s liability 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

13 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
finding is based. That is, MDalgorithms argues that because La Canada’s marks were weak and 
the field was crowded with other MD-formative marks, no reasonable jury could have concluded it 
intended to copy La Canada’s marks. 
True or not, that ignores the possibility that MDalgorithms willfully blinded itself to the 
infringement. Willful blindness contains “two basic requirements: (1) The defendant must 
subjectively believe that there is a high probability that a fact exists and (2) the defendant must 
take deliberate actions to avoid learning of that fact.” Glob.-Tech Appliances, Inc. v. SEB S.A., 563 
U.S. 754, 769 (2011). To demonstrate the first element, La Canada points to, among other 
evidence, MDalgorithms’ attempts to purchase the domain mdhair.com. If that domain was in use, 
MDalgorithms must have known that there was a “high probability,” id., that someone was 
marketing a product under that trademark—or at least a jury could so conclude. Despite that, 
MDalgorithms’ founder, Dr. Harth, admitted that he took no steps to learn who that someone was. 
See Tr. 332:4–6 (“Q: Did you ever come to learn who the owner of the mdhair.com domain name 
is? A. No.”). The jury was entitled to infer that his failure to learn such a basic fact was due to his 
desire to stay ignorant. The jury’s willfulness finding was therefore reasonable. 
C. The Alleged Reverse Confusion Theory 
MDalgorithms’ next argument in favor of judgment as a matter of law and a new trial is 
that La Canada improperly framed this as a reverse confusion case. In a case of reverse confusion, 
a smaller senior user sues a larger junior user on the theory that the latter’s infringement confused 
customers into thinking that the former’s products come from the latter. Return briefly to the 
hypothetical case of Hermès v. Hermène, but now imagine that Hermène entered the market first 
only for Hermès to show up later and win the hearts of luxury-goods consumers the world over. In 
that case, Hermène could likely state a strong reverse confusion claim because Hermès’s brand 
power and its infringement might reasonably, but mistakenly, convince consumers that Hermène’s 
products actually come from Hermès. See Surfvivor Media, Inc. v. Survivor Prods., 406 F.3d 625, 
630 (9th Cir. 2005) (“[R]everse confusion occurs when consumers dealing with the senior mark 
holder believe that they are doing business with the junior one.”). While some smaller senior users 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

14 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
might consider themselves lucky to be associated with a prestige brand like Hermès, others may 
lament the loss of the unique goodwill they tried to cultivate over years in the market. Here, 
reverse confusion would mean that, as a result of MDalgorithms’ infringement, La Canada’s 
customers thought they were dealing with MDalgorithms. La Canada previously tried to advance a 
reverse confusion theory, but that advance was rebuffed because La Canada did not plead reverse 
confusion in its complaint. 
Without question, some of the evidence and arguments at trial would have been more at 
home in a reverse confusion case. For example, Ryan Nguyen testified that this was a “David and 
Goliath situation.” Tr. 510:23–511:11. In a forward confusion case, the senior user (La Canada) is 
typically the “Goliath”: It is attempting to stop the plucky junior user from misappropriating its 
goodwill. In a reverse confusion case, by contrast, the senior user is the “David”: It is attempting 
to protect its goodwill from being overwhelmed by the market power of the junior user (the 
Goliath). Yet, puzzlingly, Nguyen testified that in his metaphor, La Canada was the David, 
indicating he thought of this as a reverse confusion case. See Tr. 524:8–10. 
La Canada also elicited testimony about the magnitude of MDalgorithms’ advertising 
spend in an attempt to persuade the jury that MDalgorithms had “flooded the market” with its 
infringing mark. See Tr. 510:23–511:11 (emphasizing that MDalgorithms spent over $9.2 million 
to market the products at issue); 725:17–20 (La Canada’s closing argument: “What are you 
supposed to do? How are you supposed to compete with the $9.2 million -- it’s not even their 
money really -- that they’re using to flood the market here with an infringing mark.”). Flooding 
the market is, however, precisely what one might expect a reverse confusion plaintiff to allege: 
that a junior user aggressively marketed its product to become the dominant player, causing 
consumers to believe that the senior user’s product actually came from the junior user. Aggressive 
advertising spend is of dubious relevance in a forward confusion case, where the whole theory is 
that the junior user is drafting off the goodwill of the senior user. 
However inadvisable these remarks were, they do not compel vacatur of the jury verdict. 
Most importantly, La Canada did not present any evidence of reverse confusion. Their central 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

15 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
evidence of actual confusion—purported MDalgorithms customers reaching out to La Canada 
about MDhair—was of actual forward confusion. Dr. Lin also testified that she had spent many 
years building the goodwill of her business and was frustrated that customers came to believe that 
she produced MDhair. See Tr. 163:19–164:10 (discussing the awards La Canada has won); 177:2–
4 (expressing frustration that an unhappy MDhair customer was contacting La Canada about being 
overcharged). 
Any residual confusion about La Canada’s theory of the case was remedied by the jury 
instructions, which explained to the jury that it “must consider whether a reasonably prudent 
consumer in the marketplace is likely to be confused as to the source of the Defendant’s hair care 
product.” Instruction 19 (emphasis added); see also Instruction 20 (“[O]ne factor to consider in 
determining whether the MDhair trademark used by the Defendant is likely to cause confusion 
with the Plaintiff’s trademarks as to the source of the Defendant’s hair care products is the 
strength of the Plaintiff’s mark.”) (emphasis added). There was no space for the jury to find 
liability on a reverse confusion theory, so the verdict need not be set aside. 
D. Unclean Hands 
Finally, MDalgorithms asks that the jury verdict be vacated because La Canada acted with 
unclean hands. “Unclean hands is a defense to a Lanham Act infringement suit.” Fuddruckers, Inc. 
v. Doc's B.R. Others, Inc., 826 F.2d 837, 847 (9th Cir.1987). “To make out an unclean hands 
defense, a trademark defendant ‘must demonstrate that the plaintiff’s conduct is inequitable and 
that the conduct relates to the subject matter of its claims.’” Japan Telecom, Inc. v. Japan Telecom 
Am. Inc., 287 F.3d 866, 870 (9th Cir. 2002) (quoting Fuddruckers, 826 F.2d at 847). “To show 
that a trademark plaintiff’s conduct is inequitable, defendant must show that plaintiff used the 
trademark to deceive consumers.” Id.; see Dollar Sys., Inc. v. Avcar Leasing Sys., Inc., 890 F.2d 
165, 173 (9th Cir.1989) (“Bad intent is the essence of the defense of unclean hands.”). Unclean 
hands is an equitable defense for the court to decide. 
MDalgorithms contends that La Canada engineered the confusion that served as the basis 
of its lawsuit. It complains specifically that La Canada purchased and redirected two domains—

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

16 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
mdhair.com and mdacne.net—to La Canada’s main website. The majority of the instances of 
actual confusion presented by La Canada—68%, MDalgorithms says—occurred while these 
redirects were occurring. 
This argument is unpersuasive. First, the relevance of La Canada’s choice to obtain and 
redirect the domain mdacne.net is unclear. The instances of actual confusion that permitted an 
inference of a likelihood of confusion were all related to MDhair: customers discussing a blue 
bottle, naming MDhair specifically, describing the solution, or the like. This case is not about 
MDacne. 
Second, MDalgorithms has failed to show that La Canada’s redirection of mdhair.com to 
its website was somehow inequitable. La Canada markets a product called MD HAIR, and it 
began to do so well before MDalgorithms entered the market. Mdhair.com is a perfectly natural 
domain from which to offer that product. See Tr. 601:3–5 (“If it’s a brand they know, there’s often 
an assumption that there’s an associated domain. And that domain name is typically .com.”). That 
MDalgorithms may have also wished to market its product at that domain (and that it would have 
made sense for it to do so) does not render La Canada’s decision to do so unfair. If anything, La 
Canada’s earlier-in-time development of its product gives it an equitable entitlement to the 
domain. 
Third, La Canada’s assertion that 68% of the instances of actual confusion occurred during 
the period in which these redirects were occurring proves far too little. Dr. Lin testified that she 
activated MDhair.com and began redirects to La Canada’s main website sometime in 2023, after 
she learned that MDalgorithms was selling its product at mdhair.co in April of that year. See Tr. 
205:10–21. Cassandra Havea testified that she first received an email from an MDhair customer in 
April 2022, see Tr. 270:8–13, and her log of confused customers goes through January 2026. 
Therefore, the redirects were up and running for between roughly 55% of the period in which 
confusion was being tracked (if we assume the redirects started in April 2023, as soon as Dr. Lin 
learned about mdhair.co) and roughly 73% of the period in which confusion was being tracked (if 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

17 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
we assume the redirects started at the end of 2023).3 It is thus entirely unsurprising that 68% of the 
instances of actual confusion occurred during the period in which the redirects were occurring. 
That is roughly what would be expected if the redirects had no impact on confusion at all. 
In sum, the jury’s liability finding was reasonable in light of the evidence, so 
MDalgorithms’ Rule 50(b) motion is denied as to liability. MDalgorithms’ motion for a new trial 
is denied as to liability for the same substantive reasons. Though evidence may be reweighed and 
credibility assessed in adjudicating a motion for a new trial, Rule 59 is not a license for the district 
court to substitute its view for the jury’s. See Roy v. Volkswagen of America, Inc., 896 F.2d 1174, 
1176 (“While the trial court may weigh the evidence and credibility of the witnesses, the court is 
not justified in granting a new trial “merely because it might have come to a different result from 
that reached by the jury.” (quoting Wilhelm v. Associated Container Transportation (Australia) 
Ltd., 648 F.2d 1197, 1198 (9th Cir. 1981))). Here, there was certainly evidence that could have 
supported a finding of no liability, but the jury’s liability verdict was nonetheless based on 
credible evidence (even accounting for the portion of Cassandra Havea’s testimony that ought to 
have been disregarded). The jury’s decision to credit particular evidence and make particular 
inferences did not work a manifest injustice on MDalgorithms, so a new trial on liability is not 
necessary. 
ii. The Jury’s Damages Award 
If the jury’s liability verdict is not set aside, MDalgorithms requests that the $1.9 million 
damages award be scaled back. Though the jury did not say so explicitly—because the verdict 
form did not require it to—the damages award was largely based on a theory of corrective 
advertising. That theory posits that a trademark infringement plaintiff should be compensated in 
 
3 The math goes like this: there are 45 months between April 2022, and January 2026—the range 
in which instances of confusion were recorded. There are 33 months between April 2023, and 
January 2026—the longest plausible period during which the redirects occurred. There are 25 
months between December 2023, and January 2026—the shortest plausible period during which 
the redirects occurred. Thus, the percentage of the period in which instances of confusion were 
recorded where redirects were also occurring is between 25/45 (55%) and 33/45 (73%). 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

18 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
the amount it will be required to spend on advertising to undo the confusion created by the 
defendant’s infringement. The award here must have been based on corrective advertising because 
the evidence did not permit a finding that La Canada suffered an injury comparable in size through 
lost profits or the like. In fact, La Canada’s damages expert, Ryan Nguyen, testified that 
MDalgorithms’ infringement cost La Canada a modest $33,205 in lost profits. See Tr. 528:13–25; 
528:18–20. 
MDalgorithms’ first objection to this corrective advertising award is trivial. It contends 
that such an award was not permitted by the jury instructions, which limited prospective costs, 
including corrective advertising, to “actual damage to the value of the Plaintiff’s mark at the time 
of the infringement by the Defendant.” Instruction 23. MDalgorithms’ argument, it would seem, 
limits the term “actual damage” to only the retrospective monetary damage that La Canada 
suffered. That makes no sense. The portion of instruction 23 that MDalgorithms references deals 
with future costs, and it is nonsensical to say that future costs cannot exceed past economic harm. 
The better reading is that the instruction prevented the jury from awarding future costs that were 
meant to overcompensate by, for example, punishing MDalgorithms—as those types of damages 
are beyond the ambit of “actual damages” as that phrase is typically understood. 
Indeed, corrective advertising is a measure of compensatory damages, making them 
“actual damage[s].” See Adray v. Adry-Mart, Inc., 76 F.3d 984, 989 (9th Cir. 1995), as amended 
on denial of reh’g (Feb. 15, 1996) (describing corrective advertising as “essentially compensatory 
damages”). That is because the goal is to give the plaintiff the means to put itself back in the 
position it would have been in absent the violation. 
MDalgorithms’ better objection is that the quantum of corrective advertising the jury 
awarded was not supported by the evidence. The most direct evidence came out during the 
testimony of Dr. Lin, who said that she had a plan to correct the confusion through advertising 
over four years. See Tr. 194:7–13 (“I’m going to spend the next four years to dig my way out of 
the hole.”). She testified that “on average” over the last four years, she (meaning La Canada) spent 
“about $300,000” on advertising. Tr. 194:15–21. 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

19 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
She was less certain about how much would be necessary going forward to remedy the 
confusion. At first, she said her plan was to “spend at least as much as what I normally spend, and 
if that’s not enough . . . spend more.” Id. She then quickly upped the ante, claiming that she 
“know[s] at least [she] need[s] to double it,” and that if that was not enough to create a response in 
the market, she would “need to go up” again. Tr. 194:23–24. If La Canada did double its 
advertising spend, the corrective advertising damages award would be $1,200,000—$300,000 
extra in spend per year multiplied by four years. See Tr. 194:25–195:3. 
The only other witness that discussed corrective advertising was Ryan Nguyen. Instead of 
focusing on La Canada’s advertising spend, Nguyen focused on MDalgorithms’ advertising spend 
on the theory that it is reasonable to calculate how much a plaintiff will have to spend to correct 
confusion by assessing how much the defendant spent creating the confusion. He testified, based 
on an assessment of MDalgorithms’ financial documents, that MDalgorithms spent $9.2 million 
on advertising between 2022 and June 2025. He then multiplied that number by 25 percent to 
arrive at a corrective advertising figure of $2.3 million.4 
The testimony of both Dr. Lin and Nguyen were deeply flawed, and no reasonable jury 
could have relied on either to deduce a non-speculative corrective advertising award. Dr. Lin’s 
testimony was pure conjecture. She took La Canada’s historical spend and said she would have to 
double it, maybe more. Based on what, exactly, one might wonder? Dr. Lin—nor any other 
witness—provided any specifics: how much confusion MDalgorithms had generated, what would 
be necessary to remedy it, and perhaps most importantly, where she came up with “double.” A 
jury is permitted to believe particular witnesses and make reasonable inferences in their favor, but 
it is not permitted to ratify guesswork. 
 
4 Nguyen also testified that the figure based on Dr. Lin’s testimony would have been $2.4 million, 
which he calculated by multiplying La Canada’s annual advertising spend, $300,000, by two 
(because Dr. Lin said she would double it) and then by four years. See Tr. 510:15–22. That math 
does not take into account La Canada’s normal rate of advertising spend. If La Canada will have to 
spend $600,000 annually but would have spent $300,000 in the normal course, MDalgorithms can 
only be liable for the additional $300,000, not the full $600,000. 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

20 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
Nguyen’s testimony fares no better. Multiplying the defendant’s historical advertising 
spend by 25 percent to estimate the quantum of corrective advertising has pedigree, but it is not 
axiomatically applied. See Big O Tire Dealers, Inc. v. Goodyear Tire & Rubber Co., 561 F.2d 
1365, 1375 (10th Cir. 1977). Nguyen made no attempt to justify the application of that standard in 
this case—there was no discussion of, for example, the pervasiveness of the confusion, the portion 
of MDalgorithms’ advertising budget that went towards MDhair, or the geographic reach of each 
company’s marketing efforts. He simply got on the stand, said “[i]n my experience, [the 25 
percent standard] does apply,” and then stepped down. That is not good enough. 
The 25 percent “rule” seems especially inappropriate here. Because corrective advertising 
is a form of compensatory damages, it must reflect the loss of goodwill or reputational damage 
that the defendant’s infringement caused. Cf. Wag Hotels, Inc. v. Wag Labs, Inc., No. 20-CV-
01326-BLF, 2023 WL 3605977, at *9 (N.D. Cal. May 22, 2023) (explaining that a plaintiff must 
present non-speculative evidence of harm to goodwill and reputation to secure a corrective 
advertising award). La Canada did not avail itself of many of the most conventional methods of 
showing diminution of goodwill. For example, it did not conduct a survey, which—while 
absolutely not required in every trademark case—might have shown that customers of hair 
regeneration products have a less favorable impression of La Canada than they did before the 
infringement. Nor did it call a marketing expert, who might have been able to testify to the likely 
brand value loss from MDalgorithms’ infringement. In the end, the only proof La Canada offered 
of loss of goodwill was loss of profits. A loss of $33,205 over three years may be something, but 
it is not much. It certainly does not establish the substantial loss of goodwill that would justify an 
award of 25 percent of the defendant’s advertising spend, especially where the defendant does an 
order of magnitude more business than does the plaintiff. 
La Canada was put on notice in the pretrial motions practice that its attempt to obtain a 
corrective advertising award would require real legwork. Though MDalgorithms’ motion in limine 
to exclude any reference to this theory was denied, it was made clear that “an infringement 
plaintiff is not automatically entitled to some percentage of defendant’s historical advertising 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

21 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
spend simply because it establishes infringement.” Dkt. 175, at 5. La Canada was also told that it 
“bears the burden of demonstrating that application of [the 25 percent standard] will result in an 
accurate approximation of what it would have to spend in the future to correct the confusion 
generated by MDalgorithms’ alleged infringement.” Id., at 6. It ought to have heeded that 
admonition. 
Ultimately, the record cannot bear the weight of the jury’s damages award. The only figure 
that is consistent with the evidence is the $33,205 in lost profits. The remainder must be set aside. 
Under Rule 50(c), a conditional ruling on MDalgorithms’ Rule 59 motion is required to the 
extent its Rule 50(b) motion is granted. Because the threshold for granting a Rule 59 motion is 
lower than that for a Rule 50(b) motion and because the arguments MDalgorithms advances in 
support of both motions are the same, the Rule 59 motion is conditionally granted in part for the 
same reasons the Rule 50 motion is granted in part. If the grant of the Rule 50(b) motion as to 
damages is subsequently reversed or vacated, a new trial on damages will follow. See Experience 
Hendrix, L.L.C. v. Hendrixlicensing.com, Ltd., No. C09-285Z, 2011 WL 4402775, at *6 (W.D. 
Wash. Sept. 21, 2011), aff'd in part, vacated in part, rev’d in part sub nom. Experience Hendrix 
L.L.C. v. Hendrixlicensing.com Ltd, 762 F.3d 829 (9th Cir. 2014). 
iii. La Canada’s Motions 
A. Permanent Injunction 
La Canada first moves for a permanent injunction to prevent MDalgorithms from 
continuing to use MDhair to market its haircare solution. See 15 U.S.C. § 1116(a). “According to 
well-established principles of equity, a plaintiff seeking a permanent injunction must satisfy a 
four-factor test before a court may grant such relief. A plaintiff must demonstrate: (1) that it has 
suffered an irreparable injury; (2) that remedies available at law, such as monetary damages, are 
inadequate to compensate for that injury; (3) that, considering the balance of hardships between 
the plaintiff and defendant, a remedy in equity is warranted; and (4) that the public interest would 
not be disserved by a permanent injunction.” eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388, 
391 (2006). 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

22 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
In the Ninth Circuit, injunctions are no longer granted reflexively upon a showing of 
trademark infringement. See Herb Reed Enters., LLC v. Fla. Ent. Mgmt., Inc., 736 F.3d 1239, 
1249 (9th Cir. 2013). Rather, “actual irreparable harm must be demonstrated to obtain a permanent 
injunction in a trademark infringement action.” Id. La Canada has failed to do so. La Canada’s 
assertion of prospective irreparable harm is basically: we were harmed in the past, the conduct is 
ongoing, therefore we will experience harm in the future. 
It is a tidy argument, but unfortunately for La Canada it has been squarely rejected by the 
Ninth Circuit. In Herb Reed, a trademark infringement plaintiff moved for a preliminary 
injunction, and the district court obliged, citing “loss of control over business reputation and 
damage to goodwill” as irreparable harm. 736 F.3d at 1250. The Ninth Circuit reversed, 
explaining that those consequences can constitute irreparable harm but that the plaintiff bears the 
burden of demonstrating that they are likely to occur absent injunctive relief. See id. There, the 
plaintiff had failed to do so because all it did was point to evidence of customer confusion. See id. 
(explaining that “an email from a potential customer complaining to Marshak’s booking agent that 
the customer wanted Herb Reed’s band rather than another tribute band . . . simply underscores 
customer confusion, not irreparable harm.”). The “practical effect” of the injunction, the panel 
thus noted, was to “reinsert the now-rejected presumption of irreparable harm based solely on a 
strong case of trademark infringement.” Id. at 1251. 
The same problem bedevils La Canada’s proffer here. It has pointed to evidence that 
supports a finding of past harm—including customer confusion and testimony about the loss of 
control over its commercial reputation—but it has not pointed to anything specific about the 
future. As in Herb Reed, its insistence that it will suffer harm in the future because it suffered 
harm in the past is “grounded in platitudes rather than evidence,” and attempts to revive a bygone 
era of trademark law. 736 F.3d at 1250 (“Gone are the days when ‘[o]nce the plaintiff in an 
infringement action has established a likelihood of confusion, it is ordinarily presumed that the 
plaintiff will suffer irreparable harm if injunctive relief does not issue.’” (quoting Rodeo 
Collection, Ltd. v. W. Seventh, 812 F.2d 1215, 1220 (9th Cir. 1987) (alteration in original))). 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

23 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
La Canada has also failed to establish that the balance of hardships supports an injunction. 
As explained, La Canada has suffered relatively minimal economic harm from this infringement, 
and it has not presented any evidence other than lost profits suggesting that its goodwill has been 
eroded. Indeed, the only burden La Canada seems sure to bear should infringement continue is 
having to respond to misdirected inquires. That pales in comparison to the cost of forcing 
MDalgorithms to rebrand one of its core products. 
La Canada’s argument as to the public interest is equally unpersuasive. It notes that some 
customers contacted La Canada regarding problems about MDhair, but it fails to explain why that 
alone presents a threat to the public interest. Those who contacted La Canada by mistake were 
rerouted to MDalgorithms, and those who were confused about source of MDhair but never 
contacted La Canada because they had nothing to complain about suffered no harm from their 
misapprehension. La Canada’s argument—that the public interest is served by an injunction 
because the jury found that the public was likely to be confused—would seem to apply to every 
trademark infringement case and thus convert a permanent injunction into a commonplace 
remedy. The motion is denied. 
B. Disgorgement of Profits 
The Lanham Act permits an award of “the defendant’s profits” consistent with “principles 
of equity.” 15 U.S.C. § 1117(a). Though willfulness is “a highly important consideration in 
determining whether an award of profits is appropriate,” it is neither necessary nor sufficient to 
obtaining such an award. Romag, 590 U.S. at 219; see Faberge, Inc. v. Saxony Prods., Inc., 605 
F.2d 426, 429 (9th Cir. 1979). Rather, the statutory directive to award disgorgement subject to 
“principles of equity” permits consideration of a broad range of factors, among them (1) diversion 
of sales; (2) adequacy of the other remedies; (3) delay by the plaintiff in asserting its trademark 
rights; (4) the public interest in deterring misconduct; (5) whether this is a case of palming off; and 
(6) whether an award of profits would give plaintiff a windfall. See Pebble Beach Co. v. Tour 18 I 
Ltd., 155 F.3d 526, 554 (5th Cir. 1998); Fifty-Six Hope Road Music, Ltd. v. A.V.E.L.A., Inc., 778 
F.3d 1059, 1073 (9th Cir. 2015) (discussing the windfall factor); Maier Brewing Co. v. 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

24 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
Fleischmann Distilling Corp., 390 F.2d 117, 121 (9th Cir. 1968) (observing that the Lanham Act 
“confers a wide scope of discretion upon the district judge in the fashioning of a remedy for a 
violation of the Act”). 
Disgorgement is not appropriate in this case. First, La Canada has provided only scant 
evidence that it competes directly with MDalgorithms, which means that very few, if any, sales 
have been diverted as a result of the infringement. La Canada attempts to establish direct 
competition through generic evidence that it has sold products under its MD-formative marks for 
almost twenty years, but that says nothing about who it has competed against during that period. 
There is no evidence that customers in the market for hair loss solutions regularly encountered 
both companies and had to decide between the two. To the contrary, MDalgorithms offered 
evidence that the two companies sold through different sales channels and offered meaningfully 
different products. While La Canada sells its products on conventional e-commerce like Amazon, 
MDalgorithms sells MDhair through a dedicated product website. See Tr. 337:8–10. Any customer 
encountering both products would have two very different experiences: to buy La Canada’s 
product a customer just has to click and pay; to buy MDalgorithms’, the bulk of customers fill out 
a survey and undergo an AI-powered image assessment before receiving a custom solution. See 
Tr. 391:21–392:22. 
Second, awarding disgorgement would give La Canada a monumental windfall. As already 
discussed, the evidence supported a damages award of $33,205, yet La Canada is asking for 
disgorgement of between $1,880,000 and $18,326,503. That range represents, on the low end, a 
57x multiplier on La Canada’s actual damages and, on the high end, a staggering 592x multiplier. 
Moreover, La Canada’s damages expert, Ryan Nguyen, testified that in 2022, La Canada made 
$147,000 in revenue, meaning that the requested disgorgement award could keep it going for 
between 12 and 125 years. That is the definition of a windfall. 
Perhaps recognizing the paucity of evidence supporting disgorgement, La Canada stakes 
much of its argument on the jury’s finding that the infringement was willful. Though 
disgorgement is ultimately an equitable issue, the jury’s factual conclusion regarding willfulness is 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

25 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
binding so long as it was reasonable. That said, willfulness exists on a spectrum, and while the 
evidence here permitted the conclusion that MDalgorithms was willfully blind to its own 
infringement, see supra section II.a.i.B, it did not permit the conclusion that MDalgorithms 
deliberately intended to capitalize on La Canada’s goodwill. La Canada nearly concedes as much, 
as the discussion of willfulness in its opposition to MDalgorithms’ motion for a new trial focuses 
exclusively on willful blindness. See Dkt. 232 at 13–16. 
That is rightly so. The only evidence La Canada offered at trial that could plausibly 
support a finding of straightforward willful infringement was (1) that its marks were conceptually 
and commercially strong (giving MDalgorithms an incentive to willfully infringe) and (2) 
correspondence between MDalgorithms and La Canada in which La Canada demanded that 
MDalgorithms cease infringing (demonstrating that MDalgorithms knew, at some point, what it 
was doing). The inference required from that evidence that MDalgorithms’ goal was to steal La 
Canada’s goodwill is simply too tenuous. Moreover, any jury inclined to make that inference 
would have had to grapple with the powerful evidence that La Canada did not have very much 
goodwill to exploit. No doubt, as acknowledged, La Canada had been in business for many years 
and had some accolades to its name by the time this dispute arose. However, in 2022—right 
around the time the infringement began—its revenues were a meager $147,000. That same year, 
MDalgorithms’ revenues were north of $10 million, and they grew to more than $17 million the 
following year. It is simply implausible that MDalgorithms would have deliberately decided to 
trade on the goodwill of a business generating only about 1.5 percent of the revenue that it was 
bringing in. That makes MDalgorithms’ conduct substantially less nefarious, weakening the case 
for disgorgement. 
The absence of misappropriated goodwill also mitigates the harm to the consuming public, 
reducing further the need for disgorgement. The Ninth Circuit has explained that the public is 
injured by trademark infringement because many consumers “are willing to pay substantial 
premiums for particular items which bear famous trademarks based on their belief that such items 
are of the same high quality as is traditionally associated with the trademark owner.” Playboy 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

26 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
Enters., Inc. v. Baccarat Clothing Co., 692 F.2d 1272, 1275 (9th Cir. 1982). When a trademark is 
infringed, those consumers are “denied the benefits of their bargains,” and a judicial remedy that 
“will take all the economic incentive out of trademark infringement” becomes necessary. Id. Here, 
by contrast, La Canada has not shown that consumers of hair regrowth products were harmed by 
MDalgorithms’ infringement. Though some consumers may have been confused about the source 
of MDhair, the evidence does not indicate that those consumers got a markedly worse product or 
that they paid a higher price for a comparable product. There is thus no particular public interest in 
crafting a remedy which will deter this conduct going forward. For all these reasons, La Canada’s 
request for disgorgement of profits is denied. 
C. Enhanced Damages 
La Canada next asks for enhanced compensatory damages. The Lanham Act permits an 
award of damages of “any sum above the amount found as actual damages, not exceeding three 
times such amount” provided that such an award “shall constitute compensation and not a 
penalty.” 15 U.S.C. § 1117(a). 
This request is denied. For reasons already explained, the jury’s damages award 
overcompensated, rather than undercompensated, La Canada—and it did so in rather spectacular 
fashion. As previously noted, the only evidence La Canada presented in support of its corrective 
advertising request were a few unsupported remarks by Dr. Lin about how much would be 
necessary to fix the confusion MDalgorithms caused and legally erroneous assertions offered by 
Ryan Nguyen about how corrective advertising works in the Ninth Circuit. The only damages 
award the evidence supported was equal to the undisputed quantum of lost profits: $33,205. 
La Canada’s request for enhanced damages must therefore be understood with this reduced 
award as the baseline. That is, the maximum that can be awarded is three times $33,205, or 
$99,615 (not $5.7 million, which is three times the original award of $1.9 million). La Canada 
points to two reasons why the jury’s damages award is insufficient to redress the harm it 
experienced. First, it says the award cannot account for the “long tail of confusion” that will occur 
after the verdict is returned. Even if the prospect of future harm could serve as the basis for 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

27 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
enhanced damages now, that prospect must be substantiated by more than just conjecture. La 
Canada’s only evidence of future harm is past harm; it has offered nothing from which to conclude 
that the problem is likely to persist. If anything, the evidence reveals the opposite. MDalgorithms 
has experienced substantial revenue growth in recent years, and as its brand power grows, it 
becomes less and less likely that its customers will buy MDhair thinking that they are purchasing a 
La Canada product. 
Second, La Canada asserts that the best estimate of funds needed for corrective advertising 
is $2.3 million, which is equal to 25 percent of MDalgorithms’ total marketing spend of $9.2 
million. Because the jury’s corrective advertising award has been vacated, this argument is a non-
starter. La Canada not only failed to demonstrate that 25 percent was an appropriate figure to use 
in this case, it failed to demonstrate the reasonableness of any other figure. La Canada continues to 
assert that Nguyen’s testimony was based on the “Ninth Circuit recognized corrective advertising 
methodology,” Dkt. 229, at 9, which, in fact, does not exist. The appropriate measure of corrective 
advertising always turns on the facts of each individual case—that 25 percent may be appropriate 
in one case does not relieve a trademark plaintiff of its obligation to prove with evidence that it is 
also appropriate in its case. La Canada was told as much as far back as the rulings on the motions 
in limine. 
D. Judgment on Partial Findings on MDalgorithms’ Counterclaims Eight and Nine 
MDalgorithms brought counterclaims for common law unfair competition (counterclaim 
eight) and for unfair competition under the California Business & Professions Code (counterclaim 
nine). Both claims are predicated on the theory that La Canada’s decision to redirect consumers 
who visited mdhair.com to its website was unfair. La Canada moves for judgment on partial 
findings on those counterclaims under Rule 52. 
To maintain a claim for unfair competition under both common law and the California 
Business & Professions Code, a plaintiff must demonstrate that it was harmed by the challenged 
conduct. See Bank of the W. v. Superior Ct., 2 Cal. 4th 1254, 1264, 833 P.2d 545, 551 (1992) 
(observing that the “common law tort of unfair competition . . . required a showing of competitive 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

28 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
injury”); Cal. Bus. & Prof. Code § 17204 (authorizing the bringing of a claim “by a person who 
has suffered injury in fact and has lost money or property as a result of unfair competition”). 
MDalgorithms conceded at trial that it suffered no injury as a result of the redirects from 
mdhair.com. See Tr. 625:25–626:4 (“Court: Okay. So you are effectively stipulating it caused no 
harm? . . . are you going to suggest at some point in this case, either an issue to me or whatever, 
that you’ve been damaged by this? A: No.”). Though the lack of any harm is the entire basis of La 
Canada’s motion, MDalgorithms fails to address it in its opposition. Accordingly, La Canada’s 
motion is granted. 
E. Attorney’s Fees 
Finally, La Canada moves for attorney’s fees. The Lanham Act permits an award of 
“reasonable attorney’s fees to the prevailing party” in “exceptional cases.” 15 U.S.C. § 1117(a). 
The “exceptional cases” requirement is construed narrowly. See Classic Media, Inc. v. Mewborn, 
532 F.3d 978, 990 (9th Cir. 2008). An “exceptional case . . . stands out from others with respect to 
the substantive strength of a party’s litigating position (considering both the governing law and the 
facts of the case) or the unreasonable manner in which the case was litigated.” SunEarth, Inc. v. 
Sun Earth Solar Power Co., 839 F.3d 1179, 1180 (9th Cir. 2016) (en banc) (quoting Octane 
Fitness, LLC v. ICON Health & Fitness, Inc., 572 U.S. 545, 554 (2014)). Though fee requests are 
ultimately evaluated using the totality of the circumstances, the Ninth Circuit—drawing on 
Supreme Court precedent—has identified a list of non-exhaustive factors to be considered, 
including “frivolousness, motivation, objective unreasonableness (both in the factual and legal 
components of the case) and the need in particular circumstances to advance considerations of 
compensation and deterrence.” Id. (quoting Octane Fitness, 572 U.S. at 1756 n.6). 
This is not an exceptional case. First, MDalgorithms’ litigating position was substantively 
strong. Though MDalgorithms did not prevail before the jury, it well could have. The jury could 
have reasonably concluded that La Canada’s evidence of actual confusion was de minimis and that 
its marks were conceptually and commercially weak. That MDalgorithms succeeded in largely 
gutting the verdict after trial reveals that its position on damages was meritorious as well. 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

29 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
Nor did MDalgorithms litigate this case in an unreasonable manner. MDalgorithms 
participated diligently in discovery and all other pre-trial matters, competently litigated the case 
during trial, and has moved for post-trial relief in good faith. La Canada’s argument to the contrary 
focuses on MDalgorithms’ assertion of various affirmative defenses and counterclaims, among 
them that Dr. Lin committed fraud on the USPTO and that La Canada acted with unclean hands. 
That argument is unpersuasive for at least three reasons. First, to the extent La Canada’s 
contention relies on the failure of MDalgorithms’ counterclaims and defenses before the jury, La 
Canada conflates substantive strength with strategic unreasonableness. A claim is not frivolous 
simply because it fails. Second, MDalgorithms was not obligated to put on witnesses to 
substantiate all of its counterclaims and defenses simply because it raised them before trial, and La 
Canada mistakes its calculated choice not to do so for vexatious strategy. After La Canada rested 
its case, MDalgorithms moved for judgement as a matter of law under Rule 50(a)—and it almost 
won. At that point, MDalgorithms may have reasonably thought that it would have been more 
effective to emphasize, through its own witnesses, the holes in La Canada’s case than to spend 
time building its own counterclaims. A party can, and often must, pivot its strategy as a trial 
proceeds in response to its perception of the other side’s case. Third, in emphasizing 
MDalgorithms’ concession that it suffered no competitive harm from La Canada’s domain 
redirects, La Canada demonstrates a misconstruction of the value of that evidence. MDalgorithms 
put on Peter Kent as a search engine optimization expert to explain the mechanism and likely 
effect of the redirects. Even after MDalgorithms had effectively dropped its counterclaims for 
unfair competition (notwithstanding its present opposition to La Canada’s motion for judgment on 
partial findings as to those counterclaims), MDalgorithms sought to leverage that testimony to 
prove its equitable defense of unclean hands. In sum, MDalgorithms did not litigate this case in an 
unreasonable manner. The case is far from exceptional, and an award of attorney’s fees is not 
warranted. 
III. CONCLUSION 
For the foregoing reasons, the post-trial motions are resolved as follows. MDalgorithms’ 

 
ORDER GRANTING IN PART AND DENYING IN PART POST-TRIAL MOTIONS 

30 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
United States District Court 
Northern District of California 
motion for judgment as a matter of law is granted in part and denied in part. The jury’s liability 
verdict is undisturbed, but its damages award is reduced to $33,205. MDalgorithms’ motion for a 
new trial is denied as to liability and conditionally granted as to damages. La Canada’s motions for 
a permanent injunction, for disgorgement, for enhanced damages, and for attorney’s fees are 
denied. Its motion for judgment on partial findings regarding MDalgorithms’ eighth and ninth 
counterclaims is granted. 
 
 
IT IS SO ORDERED. 
 
Dated: June 9, 2026 
______________________________________ 
RICHARD SEEBORG 
Chief United States District Judge 
 
 

Passage view · GavelSight