The pixel is now the liability
Regulators on both sides of the Atlantic spent the same week making the same point. The measurement layer is not a neutral utility, and in health it never was.
By Katie Delaney · 2026-08-03 · 13 min read
The FTC just sued over the measurement stack

A fox does not judge a field by its grass. It reads the trails worn through it, because the trails say who has been moving and where. Regulators have started reading healthcare marketing the same way, and the trails they are following are the ones your tags leave.
On 29 July 2026 the Federal Trade Commission, joined by Utah and by California through Los Angeles County Counsel, acted against Hims and Hers over what it calls deceptive and unlawful privacy practices. The complaint was filed in the United States District Court for the Northern District of California, and the Commission vote to authorise staff to file it was 2-0.
Our response to the lawsuit filed today by the Federal Trade Commission (FTC): This lawsuit disregards substantial evidence we provided the FTC during its nearly three-year investigation, ignores established state laws and industry standards in telehealth, and contorts the law...
For a healthcare marketing team the statutes stacked up in one filing are the part worth studying. The FTC pleads the FTC Act and the Restore Online Shoppers' Confidence Act; California pleads its False Advertising Law and, separately, its Unfair Competition Law; Utah pleads its Consumer Sales Practices Act. Five statutes, nine counts, one acquisition programme, one court.
Two platforms are named, and only two
The release names Meta and Snap, and then says "and other third parties". The complaint itself is where the real scope sits: it pleads that beyond Meta and Snap, pixels from at least Microsoft, Google, Criteo, MediaBids, PartnerCentric, PebblePost, Pinterest, Podsights (now Spotify Ad Analytics), Reddit, StackAdapt, TikTok, The Trade Desk and X were placed on the website. Fifteen advertising platforms, not two.
That distinction matters because it changes who should be worried. A telehealth brand reading the headline and checking only its Meta and Snap tags would close a fraction of the exposure the FTC actually pleads. The press release and the FTC case page are the documents of record, with TechCrunch and The Register carrying the news of record.
Two mechanisms, both of them routine
The uncomfortable thing about this complaint is how unexotic the conduct is for healthcare marketing. Nothing described is a rogue engineer or a shadow vendor. It is the standard playbook for patient acquisition marketing, run at a category where the standard playbook does not fit.
| Mechanism | What it does | Why it is different in health |
|---|---|---|
| Customer list sharing | Uploads lists of certain customers to Meta and Snap for matching and audience building | The list itself discloses that a named person is a customer of a health service |
| Third-party tracking technologies | Automatically shares events, the actions of visitors on the website | A page view on a condition-specific page is itself a health inference |
That second row is the one to sit with. In most categories a page view tells a platform someone likes trainers. On a condition-specific page it tells a platform something a person may not have told their family. The event payload does not need to contain a diagnosis to function as one.
The lesson for healthcare marketing is not that measurement is forbidden. It is that in healthcare advertising the page a tag sits on carries meaning the tag was never designed to handle, and consent language written for retail does not carry that weight.
The other half of the case: getting out
Tracking is not even the lead allegation. The FTC's release opens on billing: it alleges the company failed to disclose clearly that it charges consumers for prescriptions almost immediately after they submit an intake form, without the provider consultation customers were led to expect. Privacy is the second strand and cancellation the third.
So there are three strands, not one, and the third is about how hard it was to stop paying. Retention design is now regulated design.
According to the FTC's release, before 2023 most consumers could cancel only by contacting customer service via phone, email or chat. After online cancellation arrived in 2023, the button appeared only once a consumer selected an option to add or remove items from their order, and they then navigated several steps before they even saw the word "cancel". A cancel control hidden behind an amend-order control is the definition of a dark pattern, and the Restore Online Shoppers' Confidence Act exists for exactly this.
Findable, technically
Cancellation exists but sits behind a control labelled for something else, so the path is discoverable only by a customer willing to guess that amending an order is where stopping one lives.
Symmetrical
Cancellation is reachable in the same number of steps as sign-up, labelled with the word cancel, and available through the channel the subscription was bought in. Retention comes from the product, not the maze.
The symmetry test is the one to write into a brief: whatever it took to start, stopping should take no more. It is easy to audit, hard to argue with, and it removes an entire category of risk from a growth roadmap at the cost of a few points of involuntary retention that were never really yours.
The UK ran the same week, with weight-loss ads
On the same day, 29 July 2026, the Advertising Standards Authority upheld two rulings against Bolt Healthcare Ltd trading as Bolt Pharmacy, and the reasoning travels well beyond weight loss.
In ruling A26-1333536, breaching CAP Code rule 12.12, the ASA found two separate routes to the same breach. One, seen 16 February 2026, read "Affordable Weight Loss Starts Here" and "Price drop alert. We've cut the cost of our weight loss solution", and its reference to "prescription treatments" was itself held to promote prescription-only medicines to the public.
Another, seen 3 March 2026, read "SYSTEM CRASHED PRICING ERROR. Our system is still showing LAST YEAR'S WEIGHT LOSS PRICES. Take the 2-Minute Quiz Before It's Fixed". Separately, the landing pages compared Wegovy and Mounjaro, both prescription-only medicines. Neither an age gate on the first ad nor a filtering page on the second saved them, and the ASA told the advertiser not to promote prescription-only medicines to the public in future.
The second ruling, A26-1335405, breached rules 2.3 and 12.12. Two Facebook ads seen on 1 April 2026 were posted by an account called "UK Weight Loss Reviews" which was in fact operated by Bolt Pharmacy, carrying the claim "UK's best-price weight loss treatment". An advertiser reviewing itself under a neutral name is a straightforward failure to make clear that a marketing communication is a marketing communication.
That lag is the real operational problem for healthcare marketing. By the time a ruling publishes, the campaign has been rebuilt and retired, and the team that made the decision may have moved on. Governance has to happen at brief stage, because it certainly will not happen at ruling stage.
The standing guidance is not hidden. The ASA's advice on weight control and prescription-only medicines, its updated enforcement notice, the rulings index and the parallel General Pharmaceutical Council notice all pre-date these ads.
And a third regulator, on claims
For completeness, the FDA published a warning letter to VasoCorp on 28 July 2026, dated 9 July 2026, over website copy for two nerve-support supplements. The letter quotes a claim that the formula is "clinically proven to reduce the overall symptoms of occasional burning, tingling, and numbness by ~76%", and concludes the products are unapproved new drugs, citing sections 201(g)(1)(B), 201(p), 301(a), 301(d), 502(f)(1), 503(b)(1)(A) and 505(a). The FDA reviewed the website in March 2026 and gave 15 working days to respond, with seizure and injunction named as possible consequences. The warning letters index confirms the posting date, and the FTC's health products compliance guidance sets the parallel substantiation standard.
Five moves for a defensible healthcare marketing stack
None of this requires abandoning measurement. It requires knowing what each tag can see and proving that you decided it on purpose.
Produce a grid of which tracking technologies fire on which URLs. Most healthcare marketing teams have never seen this grid, and the condition-specific pages are where the exposure sits.
Remove third-party advertising tags from pages that reveal a condition, a symptom or a treatment. Measure those pages server-side or with first-party analytics instead.
A customer list from a health service discloses health status by existing. If audience building depends on it, the audience strategy needs rebuilding, not better consent copy.
Compliance review covers the landing page and every page reachable from the call to action. The Bolt rulings turned on destinations while the ad copy named nothing.
Audit the exit path against the entry path. Same channel, same number of steps, labelled with the word cancel. This closes the second half of the FTC complaint entirely.
In every other category a pixel measures interest. In health it records a disclosure the customer never made out loud.
The move most teams resist is the third. Customer list uploads are load-bearing in a lot of healthcare marketing strategy, and pulling them out costs measurable efficiency in the short term. It also removes the single mechanism the FTC named first, which is a trade most boards will make once it is put to them plainly.
That rebuilding work is the substance of folkfox healthcare marketing, and it leans on the measurement discipline behind SEO and GEO when paid audiences have to shrink.
What to watch once the tags come off
Stripping tags from condition pages will make healthcare marketing reporting look worse before it looks honest. Plan for that, and tell finance in advance, because a metric that falls without warning gets a programme cancelled.
Tagged condition pages
Third-party ad tags firing on condition-specific URLs. Zero is the only defensible number.
Exit steps
Clicks to cancel. Should match clicks to subscribe, whatever that number is.
Reviewed destinations
Share of live ads whose landing page passed compliance review, not just the ad copy.
Substantiated claims
Share of efficacy claims with a cited study on file before the ad went live.
Watch reviewed destinations first, because it is the cheapest to fix and it closes the exact gap both ASA rulings turned on. A destination review adds a day to a launch. A ruling adds a year to a legal file.
The number to take to the board
Report the count of third-party advertising tags firing on condition-specific pages. It is one number, it is countable this afternoon with a browser and patience, and it maps directly onto the mechanism the FTC named. A telehealth brand that cannot state it does not have a privacy programme, it has a privacy policy, and those are not the same object.
Set the baseline before you strip anything, the way a fox tests a stream with one paw rather than trusting the surface. Record current cost per acquisition, conversion rate and attributed revenue for the affected pages, then remove the tags and watch which of the three actually moves. Most teams discover the condition pages were never the ones doing the converting.
The brands that handle this well will treat it as a healthcare marketing positioning opportunity rather than a retreat. Being the telehealth service that can explain exactly what it does not collect is a genuine differentiator in a category built on discretion, and that argument belongs in brand strategy as much as in content marketing.
Frequently asked questions
What is the FTC case against Hims and Hers actually about?
Two things. First, sharing customer lists and website event data with Meta and Snap through tracking technologies. Second, subscription cancellation practices under the Restore Online Shoppers' Confidence Act. The complaint was filed on 29 July 2026 in the Northern District of California, joined by Utah and Los Angeles County.
Are tracking pixels illegal on healthcare websites?
Not inherently. The risk arises when a tag fires on a page that reveals a condition, symptom or treatment, because the event then functions as a health disclosure regardless of what the payload contains. The safer pattern is server-side or first-party measurement on those specific pages.
Why did the ASA ban ads that never named a medicine?
Because the landing pages did. In ruling A26-1333536 the ad copy referred only to affordable weight loss, but the destination compared Wegovy and Mounjaro, both prescription-only medicines. The ASA treated the whole journey as the advertisement and upheld a breach of CAP Code rule 12.12.
Can a brand run its own review account?
No. In ruling A26-1335405 the ASA found Facebook ads posted by a "UK Weight Loss Reviews" account were actually operated by the advertiser, breaching rules 2.3 and 12.12. Marketing communications must be obviously identifiable as marketing communications.
How long after a campaign runs does enforcement arrive?
In these rulings, between 119 and 163 days. Ads seen on 16 February, 3 March and 1 April 2026 were all ruled on 29 July 2026. That lag means compliance has to happen at brief stage, because the campaign will be long retired by the time a ruling publishes.
What makes a health efficacy claim safe to publish?
A cited study on file before the ad goes live, with the claim matching what the study actually measured. The FDA's warning letter to VasoCorp, posted 28 July 2026, turned on a stated 76% symptom reduction and a clinically-proven framing for a supplement, which made the product an unapproved new drug.
How EU Telehealth Providers Win AI Citations
Where telehealth growth goes once paid audiences shrink: healthcare marketing that earns the citation instead of buying the click.
Read the telehealth pieceThe platforms just made AI slop a distribution problem
The same two platforms named in this complaint, and what they now demand of the creative you run on them.
Read the paid social pieceThe promotion was approved. That was the whole problem
The parallel case in financial services: enforcement that attaches to the person who signed off the copy.
Read the approver pieceAI generated advertising just lost the right to stay quiet
The disclosure regime that landed the same week, and what it asks of regulated advertisers in the EU.
Read the disclosure piece
Measurement your legal team can read?
folkfox rebuilds healthcare marketing so growth survives scrutiny: clean tag maps, reviewed destinations, honest claims, and reporting that still works after the pixels come off.

