Medical Device Marketing After FDA Clearance: What Changes and What Does Not

WHOOP did everything the compliance playbook says to do. Its Blood Pressure Insights (BPI) feature named no disease, carried a not-for-medical-use disclaimer, and was sold as general wellness. In July 2025 the FDA found the device misbranded anyway, and gave its reason in one line: "The inefficacy of such disclaimers is demonstrated by evidence of individuals using BPI to monitor their hypertension."

If you are not sure how your own cleared claims read once a summarizer has compressed them, that is what an AI visibility audit measures.

Read that twice. The disclaimer did not fail on wording. It failed on evidence of what customers actually did with the product.

Clearance is not permission to say things. It fixes an intended use, and every claim you make afterwards is judged by how your buyers read it, not by how your legal team wrote it. Here is what actually changes at clearance, and what does not.

Key takeaways

• FDA clearance fixes an intended use. It is not approval of your messaging.
• Two regulators read your ads. FDA judges labeling against the cleared use; the FTC judges advertising for deception.
• A disclaimer is a sentence, not a control. Evidence of how people use your product can override it.
• Software and AI features carry intent through what they output, so "wellness" wording does not exempt them.
• The fix is a written claim set with a named owner, not a longer review queue.

What does FDA clearance actually grant you?

Clearance fixes one thing: the intended use. That single fact reshapes every medical device marketing decision that follows it. Most medical device marketing strategies are built around getting to clearance, and then quietly assume the hard part is over. Everything your marketing says afterwards is measured against that boundary, and the boundary was set by your submission rather than by your campaign. That's why "is this claim approved?" is the wrong question. No claim is approved. The intended use is cleared, and each claim either sits inside it or creates a new one.

The floor is concrete and worth knowing by heart. FDA's General Controls make a restricted device misbranded when its manufacturer, packer or distributor fails to carry a specific set of items in every advertisement:

"a true statement of the device's established name, prominently printed, and a brief statement of the intended uses of the devices and relevant warnings, precautions, side effects, and contradictions [sic]"

Notice what that list is and what it is not. It's a floor, and clearing it only stops you being misbranded on a technicality. It says nothing about the harder problem, which is what your claims imply once a reader has finished the page. Teams routinely pass this bar and fail the real test, because the established name and the warnings can all be present while the surrounding copy quietly asserts a use nobody cleared. The statute sets a second and much broader test right beside the first one: a restricted device is also misbranded if "its advertising is false or misleading in any particular" (21 U.S.C. 352). That is the test a checklist cannot answer, because "misleading" is judged on the whole impression rather than the presence of required elements. So a legal review that only confirms the elements are there gives false comfort. The elements are necessary. They were never sufficient.

How does a marketing claim create a new intended use?

By being read. The WHOOP Blood Pressure Insights letter of 14 July 2025 is the clearest recent example, because the company did almost everything the standard advice recommends. It avoided naming a disease, added a disclaimer, and framed the feature as general wellness. FDA found the device misbranded anyway.

The reasoning matters more than the outcome, and the letter states the rule marketers actually need:

"An article's intended use is based on the objective intent of the persons responsible for its labeling, and may be shown by the person's expressions, the design or composition of the article, or the circumstances surrounding its distribution (21 CFR 801.4)."

(the WHOOP letter) Read that as three separate exposures. What you say is only the first. The design of the feature is the second, and FDA leaned on it here, noting that the product "outputs a blood pressure measurement to users and provides the reading on a gauge that uses green, yellow, and orange color-coding to indicate a target blood pressure range." The third is the circumstances of distribution, which is where your audience lands. Careful wording was never the deciding factor. Your product design and your readers' behavior are compliance inputs, not just roadmap and marketing metrics.

There is a limit to that, and it is worth knowing precisely, because the risk is easy to overstate. The regulation carves out a safe harbor for cleared devices: "a firm would not be regarded as intending an unapproved new use for a device approved, cleared, granted marketing authorization, or exempted from premarket notification based solely on that firm's knowledge that such device was being prescribed or used by health care providers for such use." (21 CFR 801.4) Read the load-bearing word, which is "solely". Knowing isn't promoting. Knowing that clinicians use your cleared device off-label does not, on its own, put you in breach. Promoting it for that use does. So the line is not "never learn how your product gets used". The line is that the moment your marketing starts reflecting that off-label use back at the market, you have adopted it, and the safe harbor closes behind you.

Sōvyn runs this check for medtech and health-tech teams: what the engines actually say about your device, which sources they pull from, and where the qualifier gets dropped. The audit is where it starts.

What actually happens when FDA decides a claim crossed the line?

You get a short clock and a named list of consequences. The WHOOP letter sets both out exactly, and it reads better as an operational timeline than as a legal document, because the demands it makes land on marketing and engineering calendars rather than on lawyers alone.

FDA asks the firm to "notify this office in writing within fifteen business days from the date you receive this letter of the specific steps your firm has taken to address the noted violations" (CMS #709755), and it wants documented corrective action that addresses systemic problems, not a single page taken down. If that does not happen, the letter names what follows: "seizure, injunction, and civil money penalties." It also notes that other federal agencies may weigh compliance when awarding federal contracts. So the real cost is rarely the fine. It is fifteen business days of engineering, legal and marketing time, spent at whatever moment the letter arrives, plus a public record that procurement teams can read.

The escalation detail most teams miss is that the letter was not the first contact. The warning letter records two meetings in May 2025 and a letter from the firm on 2 June 2025 arguing the feature was not a device. The letter came on 14 July, after the firm said it would keep marketing anyway. There is usually a window between the agency noticing a claim and publishing a letter about it, and what you do inside that window decides which of the two happens. A team that can find every instance of a claim across its site, ads and sales decks inside a week is operating in that window. A team that cannot is not.

Software and AI features sit squarely inside this, and the letter closes the escape hatch teams usually reach for. The general-wellness exemption in section 520(o)(1)(B) covers software intended to maintain or encourage a healthy lifestyle and unrelated to diagnosis or treatment. FDA rejected it here because the feature "implies a causal link between a user's blood pressure measurement and wellness results" (FDA's letter), and because the underlying function was not low risk. The test is not whether your copy says "wellness". It is whether the thing your software outputs is inherently associated with diagnosing a condition. If your roadmap adds an algorithm that interprets, scores or flags anything clinical, your claim surface grows with the feature, and it grows in exactly the direction the agency is already reading. The test to run before the feature ships is short: write down what a user would reasonably believe the output tells them about their health. If that sentence implies a condition, you are inside 21 CFR 801.4 whatever the copy says.

Who actually polices your advertising after clearance, FDA or FTC?

Both, and the split is not where most teams assume. In the FTC's own words, "The FTC and the Food and Drug Administration (FDA) share jurisdiction over the marketing of dietary supplements, foods, drugs, devices, and other health-related products." The rough division is that FDA has primary responsibility for claims that appear in labeling, which includes the package, the inserts and other promotional materials at point of sale, while the FTC's remit runs to advertising.

The statute makes the second half explicit rather than implied. Under 21 U.S.C. 352, a restricted device is misbranded if "its advertising is false or misleading in any particular", and advertisements for a restricted device are "subject to the provisions of sections 52 through 55 of title 15" (the same statute). Title 15, sections 52 to 55, is the FTC Act. So the medical device statute itself points device advertising at the FTC's deception standard. A team that has built its entire review process around FDA has covered one of the two agencies reading the same page.

That last row is the specific post-clearance trap, and the FTC names it directly: "Advertisers should be careful not to mischaracterize the extent to which a product or claim has been reviewed, authorized, or approved by the FDA." (FTC guidance) The agency's own worked example is a cleared Class II device. A marketer clears an electronic belt for stimulating and strengthening healthy muscle, then runs an infomercial repeatedly calling the product "FDA Approved" beside a promise of losing two inches and ten pounds in thirty days. FTC's conclusion is that "the juxtaposition of the 'FDA Approved' reference and claims about weight loss and reduction in waist circumference gives the impression that the FDA has found the product to be effective for such dramatic effects." (the FTC's guidance) Every individual element was arguably true.

The combination wasn't.

That gap matters more in B2B medical device marketing than in consumer categories, because the reader is a committee and the claim gets forwarded, screenshotted and re-read by people who never saw the page it sat on. This is the failure mode a clearance-focused review process is structurally blind to. It checks each claim against the cleared indication and passes them one at a time. Nobody checks what the page means when a reader finishes it. If your best regulatory asset is the clearance you just earned, the temptation to put it beside your boldest commercial claim is enormous, and that adjacency is the thing being judged. The practical control is a rule about proximity, not vocabulary: any reference to FDA clearance carries, in the same breath, what was actually cleared. "Cleared for X" is a sentence your legal team will approve in seconds. "FDA Approved" as a standalone badge next to a performance promise is the sentence that costs you a year.

Why is a disclaimer not a control?

Because a disclaimer is a sentence and a control changes what can happen. The WHOOP letter settles the point for FDA: disclaimers were called ineffective against evidence of actual use. The FTC reaches the same place from the opposite direction, and its test is about placement rather than existence.

Working through an ad for nasal strips substantiated for snoring but not for sleep apnea, the guidance states that "A fine print disclosure of this fact at the bottom of the ad wouldn't be clear and conspicuous." (the same FTC guidance) What it recommends instead is a disclosure immediately beside the claim, in the same font size, in black on white. Both agencies are saying the same thing in different vocabularies. A sentence at the bottom of a page can't undo what the rest of the page set up, and it certainly can't undo how customers describe the product to each other.

A control is different. It decides, before anything ships, which claims may be made at all. In practice that means three things you can point at. First, a written claim set tied to the cleared intended use, where each claim is mapped to the evidence that supports it, which is the same discipline behind making claims in a regulated industry anywhere else in life sciences. Second, a named owner for that set, so approval is a decision by someone rather than a queue nobody finishes. Third, a record of what was published where, because when a question arrives about a page from eleven months ago, reconstructing it from memory is how a small problem becomes a long one. None of it is glamorous.

All of it is checkable, which is the one property a disclaimer doesn't have.

What should you build instead of a review queue?

Claim architecture rather than claim review. Review is a queue: everything waits, and it catches problems at the end when changing them is most expensive. Architecture decides once what may be said, writes it down, and then lets the team move at their own speed inside those lines.

The commercial timing argues for this too, and the shape of it is easy to misread. EY's Pulse of the MedTech Industry reports industry revenues reaching nearly US$600 billion, with "commercial leaders on track to achieve revenue growth of 6-7% again in 2025". Underneath that, "In 2024, there were only 61 MedTech M&A deals, compared with 184 in 2023." (the EY report) A two-thirds fall in deal count reads like a downturn until you see the other half: "value increased to US$50.5 billion in 2024, compared with $44.2 billion in 2023." (the same EY analysis) Fewer deals, more money. That is consolidation, not retreat, and it changes what a mid-size company has to do. A medical device product launch that once served as a shop window for acquirers now has to pay for itself. There are fewer acquirers to be rescued by, so more medtech marketing teams have to commercialize what they cleared rather than sell on the way to market. That makes the medical device launch strategy the company actually executes, rather than the one in the board deck, the thing that decides the year. At that point the speed at which your team can safely say something stops being a compliance detail and becomes a growth constraint, because every week a claim sits in review is a week the product is not being sold on its strongest true argument.

That control is also what protects you in the place most teams are not looking yet, which is the next section.

What happens when an AI answer summarizes your claims?

It compresses them, and the qualifier is the first thing it drops. FDA reads intended use from objective intent. The FTC reads an ad on its net impression. Both standards ask the same question a summarizer answers: what does a reasonable person walk away believing?

Sōvyn works with medtech and health-tech teams on exactly this seam, where regulated claims meet how buyers and AI engines actually read them. An AI visibility audit shows you where you stand today.

So the paragraph ChatGPT or a Google AI Overview gives a hospital buyer about your cleared device is not a marketing asset you control. It is a claim about your product, assembled out of your own material, and it will be read against the same standard a regulator would apply. Nobody at your company approved that sentence. You are still answerable for it.

This is where the claim set stops being a compliance artifact and starts being a medical device marketing strategy. A claim that carries its qualifier in the same breath survives compression, because a summarizer that keeps one sentence keeps both halves. A claim whose qualifier sits three scrolls down in a disclaimer does not survive it at all, and the summary that results is the version your buyer reads. That is the same structural reason FDA gave for rejecting the disclaimer in the first place: what the reader takes away is what counts. The mechanism has simply moved from a person skimming a page to a model condensing it.

AI visibility for health tech brands now depends on what those systems can safely lift. Your claim architecture is now your AI answer architecture. Same job, same failure mode, two audiences.

Frequently Asked Questions

Does FDA clearance mean my marketing claims are approved?

No. Clearance fixes an intended use, and your claims are measured against it afterwards. Nothing in the process approves a specific message, which is why teams that ask "was this claim cleared?" are asking a question the system does not answer.

Can a disclaimer protect a wellness claim?

Not against evidence of how people actually use the product. FDA called disclaimers ineffective in the WHOOP matter precisely because there was evidence of individuals using the feature to monitor hypertension. Wording is not the deciding factor when behavior points the other way.

What does an FDA warning letter actually demand?

A written response inside fifteen business days, with documented corrective action that fixes the system rather than one page, and a named list of what follows if it does not arrive.

Do AI and software features change what I can claim?

They already have. FDA rejected the general-wellness exemption for WHOOP because the feature implied a causal link between a measurement and wellness results, and because the underlying function was not low risk. If your software interprets, scores or flags anything clinical, that output carries intent regardless of how the copy is worded.

What has to appear in a restricted-device advertisement?

The device's established name prominently printed, plus a brief statement of intended uses and the relevant warnings, precautions, side effects, and contraindications. Treat that as the floor you clear before strategy starts, not as the strategy.

Who should own claims after clearance?

One named person, working from a written claim set. A rotating review queue spreads responsibility until nobody holds it, and that is usually what teams discover when they go looking for who approved a page.

One limit worth stating plainly

This is how we read the public record as marketers, not legal advice. It covers the two federal agencies that police claims, and it does not cover state-level rules, international regimes, or the promotional-review obligations your quality system may already impose. Every claim decision described here should end with your regulatory lead, not with us. What we are arguing is narrower and, we think, more useful: the review process most teams run is aimed at the wrong test.

Where to start this quarter

Start by writing the claim set down. This is the work we look for first when we audit medical device marketing agencies and in-house teams alike. This is one list of every claim live on your site, your sales deck and your product pages today, with each one mapped to the cleared intended use and to the evidence behind it. You'll find claims nobody can source, and those are the ones to fix first. That is a week of work for most teams, and it replaces a compliance program nobody will run.

Then add one rule to the set, because it is the cheapest risk you will remove all quarter. Every mention of your clearance travels with what was actually cleared, in the same sentence, and never sits next to a performance claim it does not cover. That single rule is what the FTC's own guidance is asking for when it warns marketers "not to mischaracterize the extent to which a product or claim has been reviewed, authorized, or approved by the FDA." (FTC's worked example) It costs nothing to apply, it survives staff turnover because it is a rule rather than a judgement, and it closes the exact gap that turned a legitimately cleared Class II device into a deceptive infomercial.

Ninety days from now you will know something you cannot answer today: which of your public claims you could defend on paper, and which you have simply been repeating. Most teams in this market cannot answer that at all.

References

1. General Controls for Medical Devices, U.S. Food and Drug Administration.
2. Warning Letter to WHOOP, Inc., CMS #709755, 14 July 2025, Office of Product Evaluation and Quality, Center for Devices and Radiological Health, U.S. Food and Drug Administration.
3. 21 CFR 801.4, Meaning of intended uses, Office of the Federal Register.
4. Health Products Compliance Guidance, U.S. Federal Trade Commission, December 2022.
5. 21 U.S.C. 352, Misbranded drugs and devices, Office of the Law Revision Counsel, U.S. House of Representatives.
6. Pulse of the MedTech Industry Report, EY.