11th Circuit affirms defense TM verdict; evidence of confusion is not evidence of harm for disgorgement

Florida Virtual School v. K12, Inc., 2026 WL 127063, No.24-10449 (11^th Cir. Jan. 15, 2026)

A pretty good example of why granting registrations tohighly descriptive (at best) terms is a bad idea! Also a good example of whynot having a harm requirement in trademark infringement encourages this kind oflitigation—although the lack of harm matters to remedies, it isn’t part of themain case, making litigation seem much more attractive.

Florida Virtual, a state-funded initiative, has federalregistrations for “Florida Virtual School” and “FLVS” for educational services.K12, a for-profit competitor, initially launched with “Florida Virtual Academy”and “Florida Virtual Program.” Florida Virtual sued K12, which settled andadopted the name “Florida Cyber Charter Academy.” But Florida Virtual suedagain when K12 launched a new program, “Florida Online School,” adding unfaircompetition, false advertising, and breach of contract claims. K12 counterclaimedfor cancellation of Florida Virtual’s registrations for fraud against theUSPTO. The trialcourt rejected all the claims (fraud on the PTO was barred by thesettlement agreement despite some decent evidence of misrepresentation), andthe court of appeals affirmed.

Of relevance to the breach of contract claims: K12 agreed to(1) pay Florida Virtual $600,000; (2) stop using the Florida Virtual Academy(FLVA) and Florida Virtual Program (FLVP) names and acronyms; (3) not useadditional “Prohibited Marks”; and (4) transfer domain names containing theprohibited marks to Florida Virtual in 2016, until when it could use them toredirect to itself. The settlement agreement included a list of “ApprovedMarks” available to K12, but the parties agreed that there would “be no presumptionagainst K12’s choice of a mark” not on that list.

As is not uncommon, nobody apparently followed up on thedomain name transfer. If you have outside counsel do the litigation, you musthave someone internal calendar issues like this for your team! Followup iswhere things are most likely to fall apart. This has been your practice pointerfor the day! Thus, when Florida Virtual objected to K12’s Florida Online School(FLOS) in 2019, it also raised concerns with K12’s continued use of FLVA.com asa redirect to its other websites. K12 then transferred the FLVA.com domain toFlorida Virtual and began the process of renaming its program “Digital Academyof Florida,” but Florida Virtual still sued.

False advertising: The false advertising claim was based ona checklist on K12’s website for “comparing K12 to other online learningsolutions.” The checklist showed two columns, each listing several features ofan online education program. “K12-Powered Schools” showed checked boxes next toeach feature while “Other Online Learning Solutions” had an unchecked box nextto each.

Florida Virtual’s survey expert concluded that the checklistmisled around 18 percent of consumers into believing that Florida Virtualoffered services that its competitors did not, but the district court concludedthat the survey portrayed the checklist “out of context” and granted summaryjudgment on the false advertising claim because there was no other evidence ofconsumer deception.

The analysis here is weird; the court didn’t like that thesurvey didn’t show parents other virtual school options and ask them if theyactually had the features at issue—but that’s falsity, which usually isestablished by evidence other than the survey. Still, the court reasoned, thechecklist wasn’t literally false, because, although one reasonable reading ofthe checklist is that K12 provided “the checked services while other schools[did] not,” another reasonable reading was that K12 was “inviting consumers todo their own research and fill out the checklist—not stating that it possessedfeatures the other providers definitely did not.” (The vagueness of the “other” category leads me to asimilar ultimate conclusion—if such a comparison is even falsifiable, itdoesn’t seem that Florida Virtual showed that all other online options in facthad the features in question. A better criticism would be that the survey didn't test the alternate meaning if it didn't give respondents the option to say "this is a checklist I can use" or something like that, along with "this means those other schools don't have those features.")

Even assuming falsity as to Florida Virtual—which did offerall the features—the survey “did not allow respondents to review the websitesof K12’s competitors and assess whether they provided the same services as K12.”[Again, this is about falsity, not misleadingness.] Thus, the survey wasunreliable for assessing a “marketing tool” whose stated purpose was to allowusers to “weigh [their] options” when comparing K12 to other providers.

Trademark infringement: The district court excluded thelost-profits testimony of Florida Virtual’s damages expert, who wrongly/withoutfoundation assumed that every Florida Online School student would have enrolledin Florida Virtual School absent the alleged infringement. With thislost-profits testimony excluded, there was no evidence of actual damages.

Florida Virtual sought disgorgement of not only K12’sprofits related to Florida Online School, but also the profits from its otherprograms because the continued use of FLVA.com as a redirect to these programs’websites was allegedly an independent act of trademark infringement. Thedistrict court disagreed and struck testimony unrelated to Florida OnlineSchool; Florida Virtual had not “based its trademark infringement arguments on[K12’s] use of the FLVA.com domain,” and the claim was released by the settlementagreement.

The court of appeals affirmed the rejection of FloridaVirtual’s actual damages remedy. The only evidence Florida Virtual had ofdamage did not show that confusion caused the damage. First, a parent testifiedthat she wanted to enroll her daughter in Florida Virtual School in 2020, butaccidentally enrolled her in Florida Online School instead. But she realizedher mistake and withdrew her daughter before classes began, then attempted toenroll her daughter with Florida Virtual, but ultimately “decided to go back tobrick-and-mortar at the end of the day” (at the point that Covid shutdowns inFlorida had ended). This was not a lost customer.

Second, there was other arguable evidence of confusion amongstudents, parents, and school officials. “But there is a difference betweengeneral confusion and actual damages, and Florida Virtual did not bridge thatgap.” The court highlighted some examples (most of which arguably just showthat the purported mark is near-generic or generic):

A social worker contacted FloridaVirtual for a Florida Online School student’s enrollment records after thestudent’s father said he had “been enrolled in FLOS (Florida Online School)which is a part of FLVS.”

A sixth grade Florida Online Schoolstudent told his teacher in an email that he was “just starting Florida VirtualSchool.”

A parent emailed her son’s FloridaOnline School teacher to withdraw him “from Florida virtual school.”

In an email to a Florida OnlineSchool teacher, a parent said, “I am new to the Florida virtual school.”

A parent contacted both FloridaVirtual and Florida Online School employees to ask about the status of herdaughter’s enrollment in Florida Online School.

Even viewed in the light most favorable to Florida Virtual, “theseexamples demonstrate confusion—but that’s all. They do not show that theconfusion diverted students from Florida Virtual to K12, or otherwise injuredFlorida Virtual.” And the damages calculation was no help because the expert assumedthat Florida Virtual would have obtained all of K12’s registrations absent theallegedly unlawful conduct. “That conclusion was not an abuse of discretion.”

But, because there’s no harm requirement, that didn’t endthe case, just kept it a bench trial.

There was no error in finding Florida Virtual’s marks weak.FV conceded descriptiveness, but the 11^th Circuit presumes relativestrength from incontestable registrations (boo). Still, that presumption can berebutted by showing commercial weakness, which K12 did. Florida Virtual’sdirector of marketing testified that it had changed its logo six times since1997 and acknowledged that changing a logo “can dilute a brand.” And its seniordirector of marketing and communications “discussed a nearly $5 million effortto rebrand [Florida Virtual’s] global operations as recently as 2020.” 

“In a 2018 survey, only 30 percent of parents withschool-aged children recognized Florida Virtual’s brand—even when prompted. Andin a 2020 survey, just 1 percent of respondents named Florida Virtual as anonline education provider without prompting.” There was other survey evidenceshowing 50% prompted recognition, but that wasn’t much more than K12’s. Therewas also evidence of third-party use of “Virtual School” modified by the namesof various Florida school districts; though FV argued that the geographicdesignation removed any confusing similarity, the district court couldreasonably take a different view.

On similarity, the word marks were “nearly identical,” but FloridaVirtual “operates in a crowded field of similar marks on similar goods orservices,” where “slight differences in names may be meaningful,” and thedesign marks looked “nothing alike.” There was no error in finding similarityto be neutral.

Customer overlap: the court found this factor neutral becauseFlorida Online School’s only customer was Hendry County School District, not“individual parents and students.” Florida Virtual argued that it “alsopartners with school districts,” so its customers are similar either way, andFlorida Online School still “catered to the same general kinds of individuals,”which was all that was required. “[W]ere we reviewing de novo, we might agreethat this factor weighs in Florida Virtual’s favor. But we are not—and it wasnot clear error for the court to determine that this factor was neutral.” K12presented evidence at trial that the Hendry County School District was the onlyone purchasing services from Florida Online School, meaning there was nooverlap. Anyway, “error in its analysis of one of the subsidiary factors” is“not enough to allow us to overturn” the trial court’s decision.

Similarity of advertising: Both parties “use[d] digitalmedia to reach their customers and facilitate services,” but they targeteddifferent audiences: K12 “primarily market[ed] to school districts,” whileFlorida Virtual advertised directly to students and parents. This minimized theoverlap.

Intent: “While there was some evidence—like K12’s continueduse of FLVA.com—that could suggest intent to infringe, other evidence supportedthe court’s finding,” including the name changes when challenged.

Actual confusion: “Short-lived confusion or confusion ofindividuals casually acquainted with a business is worthy of little weight,while confusion of actual customers of a business is worthy of substantialweight.” Reasonable minds could disagree whether it was the marks that causedany confusion reported, and thus the trial court did not clearly err.

For example, the parent mentioned above testified that shebelieved the two programs to be “one in [sic] the same,” because she “thoughtthere was only one” online education provider in Florida. Because of thatbelief, she “didn’t feel the need to research” her options “in depth.” “[I]twas reasonable for the court to conclude that the source of her confusion washer mistaken belief that there was only one provider, not the similarity ofK12’s marks. After all, if [the parent] was convinced there was only one onlineprovider, it would not make a difference to her whether the program she signedup for was called Florida Virtual School, Florida Online School, or somethingcompletely different, like Digital Academy of Florida.” The story was similarwith the other purportedly confused parent, who the district court found wasconfused about the flexibility of the schedule offered, not the name of theschool. “[S]he testified that she did not care which program her son went to,so long as it had a flexible schedule.”

It’s nice to see some focus on causation here! The courtcompared the situation to one in which a skier believes that only one airline,Delta Air Lines, offers a flight from Atlanta to Salt Lake City; she books thefirst flight that comes up in her search, which happens to be on American. “Didshe book with American instead of Delta because their names were too similar?Of course not—it’s because she thought there was only one option.” (Cf. Conopco,Inc. v. May Dept. Stores Co., 46 F.3d 1556 (Fed. Cir. 1994) (rejecting similarevidence of actual confusion where consumer testified that she believed thatnational brands made the products used in house-branded alternatives).

Florida Virtual also argued that it was error for the courtto discount evidence demonstrating actual confusion: twenty-one emails fromemployees, parents, students, and others. But it was not clear error for thedistrict court to find that the emails were not reliable evidence of confusion,but rather of “the fact that online educational service providers exist in amuddled marketplace replete with generically and descriptively namedparticipants.” It was also not clear error to point out that, without surveyevidence, there was “no way to filter out latent marketplace confusion that theparties agree exists in the online education market.”

Consumer sophistication: The trial court found that FloridaVirtual’s customers were sophisticated given “the nature and importance of aparent’s choice of where to educate their child.” This is, of course, anormative statement, not an empirical one, as the parents above demonstrated.Students looking for a college are “relatively sophisticated consumers” becauseof “the nature, importance, and size of the investment in a college education.”It was not clear error to apply that logic to schools where “parents, notstudents, are the ones making that decision. Plus, the evidence showed thatsome of Florida Virtual’s customers were school districts and administrators,and we would expect them to have a developed understanding of their onlineeducation options.”