annoyingly redacted opinion finds Block falsely advertised versus TurboTax

Intuit Inc. v. HRB Tax Gp., Inc., 2024 WL 5320392, No.5:24-cv-00253-BLF (N.D. Cal. Dec. 3, 2024)

Intuit, which makes TurboTax andhas had some advertisingtroubles of its own (hey, check out the IRS FreeFile program, for taxpayers with AGI of $84,000 or less and with many fillableforms for those with more), sued HRB (Block) for its advertising of a competingtax product, and succeeded in part.

TurboTax has three tiers: TT Do-it-yourself (DIY), LiveAssisted, and Live Full Service. Live Assisted, the focus of this case, allowscustomers to prepare their tax returns largely independently, but it alsoprovides “unlimited access to tax experts to help them with any questions thatthey may have,” among other features. (Live Full Service, by contrast, involvesactually turning over all the taxpayer’s documents to a tax professional toprepare the return.)

Within TT Live Assisted, there are multiple individualproducts that a customer may select based “upon the level of tax complexity” oftheir situation. TT Live Assisted Basic might apply to a consumer with just aW-2 and no additional schedules. TT Live Assisted Deluxe is for a “slightlymore complex” tax situation, e.g. “additional deductions and credits” or“mortgage interest.” Live Assisted Premium, “essentially includes all tax forms[and] schedules.” But all have tax expert assistance including “expert finalreview,” following a prompt asking whether the consumer wants to connect withan expert about their tax return.

Block has two tiers: Do-It-Yourself (DIY) and “File with aTax Pro.” Within DIY, there are a further four tiers, including Free Online,Deluxe, Premium, and Self-Employed. DIY Free Online is geared toward simple taxreturns and is free. DIY Deluxe is for consumers with more complex taxsituations, such as those seeking to “maximiz[e] deduction and credits,” so itincludes more tax forms. DIY Premium allows reporting of investment or rentalincome. DIY Self-Employed is for self-employed taxpayers who need to file aSchedule C.

DIY Deluxe had live expert help, artificial intelligenceassistance, an automated accuracy review or error check, year-round support,and various guarantees, among other features. Separately, Block offers “Tax ProReview” as an add-on, which allows “a DIY consumer, after they have filled outtheir taxes, the ability to send their tax return to a tax expert” along withtheir source documents, so that the tax expert can review, sign, and file thereturn.

Intuit challenged five of Block’s ad claims: (1) claims thatBlock’s products with expert and AI assistance features start at a lower pricethan Intuit’s products with similar features; (2) claims that Block’s paid DIYproducts are comparable to TurboTax Live Assisted; (3) claims that Block’s productswith AI and expert assistance cost “[a]t least $54 less than TurboTax Live”; (4)claims that Intuit’s Live Full Service product “starts at” $169; and (5) claimssuggesting that “5 million ‘TurboTired’ TurboTax consumers ‘switched’ to Blockin 2023.”

For example, Block advertised: “Fed up with hidden fees?Make the switch to H&R Block with upfront transparent pricing.” TTchallenged the middle tile’s claim that Block offered “expert help” “[s]tartingat $35” when Intuit’s “starting at” price for the same “expert help” was $0.

comparative ad

Also, relevant to claim (2), the ad compared a TurboTax LiveAssisted product—Live Assisted Deluxe—with Block’s paid DIY products that don’tinclude a final expert review, which Intuit alleged was a material difference. TTLive and Live Assisted included guarantees that Block didn’t for its paid DIYproducts. Finally, Intuit, argued that Block’s products used chatbots to makeit “incredibly difficult, if not impossible, to actually engage with anexpert,” whereas TT didn’t.

For claim (4), Intuit dropped the price of Live Full Serviceto many customers at a starting price of $89, although in prior years thestarting price had been $169; until it was contacted by Intuit regarding theinaccuracy, Block ran ads stating that the starting price for Intuit’s LiveFull Service product was $169.

For (5), Intuit pointed to an email: “It’s Better with Block”that also included “TurboTired? Switch and save with Block” and concluded:“Join the 5 million+ who switched to Block last year.”

TurboTired/join the 5+million who switched ad

Intuit offered Professor Joel Steckel of the New YorkUniversity Stern School of Business as an expert who conducted several surveys.He testified that respondents looking at Block’s home page—which displayed anadvertisement similar to the challenged ads—believed that “there were at leastas many features in the Block website and that the Block product costs less.” Basedon another study, in combination with a review of academic literature, Dr.Steckel concluded that “Block’s conduct, which resulted in the perception of anequal or superior product at a lower price, could cause negative feelingsregarding the pricing of TurboTax products and harm the TurboTax brand.” And athird study looking at the TurboTired ad led him to conclude that claims that“5 million+” people switched to using Block products “likely did influence ...consumers to switch to Block” because those consumers “would have understoodthat ... the 5 million-plus who switched to Block had all switched there fromTurboTax.”

Block’s own expert, Hal Poret, critiqued these surveys andthe related testimony, including use of controls.

Claims about expert/AI help: Intuit argued that Block“falsely claim[s] that its paid DIY products offering ‘expert’ and ‘AI’assistance ‘starting at’ $35 were cheaper than Intuit’s product with expert andAI help.” These challenged ads were “literally false,” according to Intuit,because TurboTax Live Assisted Basic—which is free for certaincustomers—includes expert and AI assistance. Block responded that the point ofthe ad was not to compare the starting price for expert/AI assistance in anycircumstance, but rather to compare Block’s DIY Deluxe product with Intuit’sTurboTax Live Assisted Deluxe product, and that the prices listed wereliterally true insofar as those were the products being compared.

The court found that Intuit’s reading of the ad was correctand that the ad was literally false by necessary implication. Block’s own witnesstestified that the bullet points listed on the relevant advertising tilepertained to various Block products, not just Block’s DIY Deluxe product: thetile mentions “[o]ptions for deductions, investors, and self-employed,” astatement that alludes to Block’s DIY Premium and DIY Self-Employed products.

Thus, it wasn’t credible that the ad was supposed to compareTurboTax Live Assisted Deluxe with Block DIY Deluxe based on “the taxsituations that th[ose] product[s] cover” and “the tax forms that areavailable” in those products. The advertising tile clearly emphasized theavailability of expert assistance—other than the $35 price, the most prominenttext on the tile states “File your taxes with expert help.” And it said nothingabout the specific tax forms available through Block’s DIY Deluxe product;instead, it alluded to multiple different Block products. “Block could easilyhave emphasized certain specific tax forms if indeed that was the basis for theprice comparison, but it did not do so. Instead, it told customers about thestarting price for Block’s products that include expert help. A viewer of thisad would readily recognize the ‘necessary implication’ that the TurboTaxproduct it listed as a comparison was the lowest-priced TurboTax product withexpert assistance.” The court therefore presumed actual deception as well.

The court also found likely success on materiality. (Here’swhere harm causation gets tricky—since Intuit doesn’t directly benefit fromanyone who uses the free product with expert assistance, and people who pay forTT because they mistakenly thought they’d get to use the free product are moreaccurately characterized as victims of false advertising rather than as beneficiaries,the “value” that Intuit gets here is actually at most “brand value” rather thanthe value of a truthful claim. We’re in a world where that’s enough, I guess.)

Steckel’s study of the ad at issue concluded that“respondents thought that there were at least as many features in the Blockwebsite and that the Block product costs less.” He also found that “price [was]extremely important,” and he also found that approximately one-third of surveyrespondents identified availability of tax expert assistance as important. Healso summarized academic literature showing that “price comparisons ...increase[ ] the purchase likelihood of the brand making the comparison andreduce[ ] consumers’ intent to search for more information.” “In combination,Dr. Steckel’s findings support his conclusion that claims that tax expertassistance is available starting at a lower price through Block’s products arelikely to affect consumer purchasing decisions, since a significant number ofconsumers are interested in expert assistance and a significant number ofconsumers also care strongly about the price of the product they select.”

Block’s critiques were insufficient; Poret primarilychallenged the lack of a proper control group, which went primarily went tomisleadingness/deception. And even without a proper control, the court wasn’t “persuadedthat it should ignore Dr. Steckel’s finding that 25 percent of respondents whoviewed the original advertisement thought it showed that Block’s productsincluded comparable or more features than the comparator product.”

The court also found likely injury to Intuit, applying apresumption of injury when there’s direct competition and a tendency to mislead,as well as testimony of Intuit’s Director of Marketing Strategy of herbelief  that the ads “absolutely wereharmful” to Intuit, in the form of both reputational harm and economic harm andSteckel’s testimony that “Block’s conduct, which resulted in the perception ofan equal or superior product at a lower price, could cause negative feelingsregarding the pricing of TurboTax products and harm the TurboTax brand” andconclusions from the academic literature that “price comparisons ... increase[] the purchase likelihood of the brand making the comparison and reduce[ ]consumers’ intent to search for more information.”

To the extent Block was enjoined from claiming that expertand AI assistance “starts at” a lower price with Block than with Intuit, Block wouldlikewise be barred from claiming that expert and AI assistance costs at least$54 less with Block than with Intuit, given that Intuit provides expert and AIassistance for free through TurboTax Live Assisted Basic, so such assistance isactually available for less with Intuit than with Block.

What about the claims comparing Block’s paid DIY productswith TT Live Assisted? Intuit identified several putative material differences:(1) although both companies’ products offer “as-you-go” expert assistance, onlyTurboTax Live Assisted has the additional “expert final review” feature withoutadditional cost; (2) this also allows a free upgrade for an expert to prepare,sign, and file the review; (3) TurboTax Live Assisted includes a guarantee thatconsumers will be reimbursed for any penalties assessed due to errors made bythe tax expert who assisted the Live Assisted consumer; (4) TT’s access to liveexperts is easier.

Block argued that comparing the two was not misleading andthat these putative differences were marketing gimmicks that weren’t significantlydifferent; the vast majority of TT Live Assisted consumers, it said, didn’t getan expert final review, since they must take proactive steps to prompt thereview process.

Lanham Act case law accepts an apples-to-oranges theory offalsity for comparative advertising when an ad “omits differences which wouldhave been material to recipients.” The court found that Intuit’s feature wasgenuinely different from Block’s on expert final review, citing some redactedevidence. Its tools “enable Intuit’s experts to proactively identify anyoverarching concerns about consumers’ prepared returns.” Although bothcompanies’ products permit unlimited question-and-answer with live experts, atthe end of the process, an Intuit customer can request a final review in whichthe expert conducting the review proactively looks for issues and uses uniquetools to help target any such issues. Block doesn’t have that “proactive”feature for its paid DIY products.

In addition, TT Live Assisted permits a tax expert to“complete, sign, and file” a consumer’s return on their behalf, while a Blockconsumer looking for a similar sign-and-file service must “pay the add-on feefor Tax Pro Review.” Although this free upgrade occurs in only a “tiny fraction”of cases (again, redacted), the court found that this was also a genuinedifference.  However, the argued ease ofaccess to expert assistance was not shown to be a genuine difference. Bothproducts included “as-you-go” assistance from experts that can be reached usinga “help button” from virtually any page of the tax preparation softwareexperience. Both had live help via web chat, phone call, or screen share. AlthoughBlock’s web chat uses an “AI Assistant” to ask an initial set of questionsbefore connecting a consumer to a live expert, the court was persuaded Block’sevidence and arguments showing that the exchange with the AI assistant wasquite brief. And although some Block experts are located in India, they receivethe same training and interact with consumers the same way regardless of theirlocation.

Finally, Intuit’s argument that the parties’ guarantees weremeaningfully different was mistaken.  

So, were expert final review and/or free sign-and-file usedin a “tiny fraction” of cases material? The court indicated that materialitycould be assessed in two ways: whether the omitted information was “ ‘likely toinfluence the purchasing decision’ of consumers,” generally demonstratedthrough use of consumer surveys, or whether “the defendants misrepresented aninherent quality or characteristic of the product.” While the “inherentqualit[ies] or characteristic[s]” of a product may sometimes beself-explanatory, any doubts as to what counts as “the very nature” of theproduct “must be addressed by evidence of why a consumer sought out a givenproduct—i.e., the primary purpose(s) that drove their consumption activity.”

Expert final review was neither an “inherent quality orcharacteristic” of Live Assisted nor a “key product feature” going to “the verynature” of an online tax preparation product, unlike the way that “the amountof beef in a burger is an inherent quality or characteristic of a burger.” Evidencethat a redacted percentage of Live Assisted consumers take advantage of theexpert final review feature didn’t help. [Hmph.] “The Court is skeptical that afeature that must be affirmatively invoked—and if not invoked, is notexperienced by the consumer—counts as part of ‘the very nature’ of a product.”There was no other evidence that the prospect of such review drove consumption.Evidence that consumers who used expert final review really liked it did notmean it was material to a purchase decision. Nor did Steckel study expert finalreview specifically. “[I]t is possible that consumers do not see a differencein value between the type of expert final review offered by Intuit and theas-you-go expert assistance offered by a live expert through either party’sproducts.” For the extraordinary remedy of a preliminary injunction, this wasinsufficient.

Similar problems hampered Intuit’s showing on the materialityof the free sign and file service, especially since Intuit doesn’t widelyadvertise this service, which requires “dig[ging] deep enough” online to evenknow that such an upgrade is available and “the fraction of individuals who areupgraded in this way is vanishingly slim.”

Claims that over five million consumers switched to Block,when only redacted [ugh!] did: Intuit argued that 5 million was the totalnumber of new Block customers. Block added a disclaimer to that effect on itswebsite after Intuit filed suit, but Intuit argued that it was insufficient.Block also argued that it discontinued any combination of this claim with the “TurboTaxswitcher” message so the claim was moot. The court disagreed, because voluntarycessation doesn’t moot a claim unless there’s no reasonable expectation thatthe wrong will be repeated. Here, Block only added a disclosure after the TROhearing and that was just an asterisk linked to small text at the bottom of thepage, not directly below the claim.

The court found likely success on falsity. The ad at issuenecessarily implied that over 5 million people switched from Intuit to Block. Thephrases “TurboTired? Switch and save with Block,” “Switch from TurboTax® now,”and “Join the 5 million+ who switched to Block last year” all appeared in thesame content box within the email. Plus, even if it weren’t false by necessary implication,Steckel’s survey showed likely deception in more than 50% of respondents; evenwith the disclaimer language in the body of the email, Steckel found that 47percent of respondents still came to the same conclusion.

Poret critiqued the survey by arguing that TurboTax andH&R Block are the “dominant names” in the field of online tax preparation,so people would just guess they’d switched from TT. But “Block has offered nocase law supporting Mr. Poret’s conclusion that a net deception measurement of ‘10to 20 percent or higher’ is necessary to show ‘enough evidence that an ad ismisleading.’” Then there are redactions that make the discussion hard tocomprehend, but seem to have something to do with Block’s intent. “In otherwords, Block itself has created relevant ‘noise’ that may lead respondents tobelieve the ‘5 million+’ switched language refers to people switching fromTurboTax specifically, regardless of whether the specific iteration of theadvertisement the respondents review says anything about TurboTax.”

Anyway, the court found that the ad was false, without needof a “net deception baseline (or a specific quantitative value thereof),” asinformed by the court’s own “experience and understanding of human nature.” Whilethe materiality burden was high, for a preliminary injunction, “Intuit has aslightly more forgiving burden for proving actual or likely deception.” Steckel’stestimony met this burden. “That a large portion of respondents were alsoconfused by Dr. Steckel’s control version of the email does not change the factthat many viewers of Block’s advertising claim came away from it with the wrongconclusion.” [It just means the control also was deceptive!]

Materiality/injury: Block’s own redacted internal materialssupported materiality. [What were they????] “Block clearly believes thatadvertising about switching to Block has in past years successfully encouragedconsumers to make such a switch.”

Claims that TT Live Full Service started at $169: Intuitargued that Full Service was available at $89 to a redacted [!!] percentage ofcustomers as “test pricing.” Block argued that its pricing team checked TT’swebsite and saw the $169 price and had no reason to believe that Intuit wouldchange its prices in January. Block contended that it couldn’t be held liable “ifIntuit is posting two different prices online” without disclosing that pricesmay differ in certain test markets.

Although false advertising is nominally strict liability,and although the strength of this argument as a matter of consumer protectionis very hard to evaluate given the redaction (if the test pricing was availableto 2% I’d feel very differently than if it was available to 75% of customers),the court was attracted to the mootness argument here to avoid the issue. Block’switness testified that a Block manager checked Intuit’s website every day tovalidate the $169 comparison price. When Block was contacted by Intuit’s teamabout the $89 price, Block “put a process in place to update that content andcorrect the mistake on [Block’s] site,” and Block represented that it does notplan to run ads using an inaccurate starting price for TurboTax Live FullService in the future. There were no circumstances “raising the specter ofgamesmanship” here; this was just a mistake that had been addressed, and thecourt denied the request for an injunction as moot.  

For the claims that survived the other factors, there was apresumption of irreparable harm, which Block didn’t rebut, and “[e]vidence ofthreatened loss of prospective customers or goodwill certainly supports afinding of the possibility of irreparable harm.” Plus there was Steckel’stestimony that “Block’s conduct, which resulted in the perception of an equalor superior product at a lower price, could cause negative feelings regardingthe pricing of TurboTax products and harm the TurboTax brand” and could“influence ... consumers to switch to Block.” Thus, Intuit was entitled to aninjunction against Block ads (1) suggesting that the “starting” price forexpert and artificial intelligence assistance is lower with Block’s productsthan with Intuit’s products, and (2) placing language about “5 million+” peoplehaving switched to Block in proximity to language about switching from TurboTaxor other language inviting the inference that all 5 million+ people switched toBlock’s products from Intuit’s products.