KLM-LOG Blog
Before You Scale Telehealth Spend for Black Friday: A 10-Point Tracking Check
November 1, 2026 · Joe · 4 min read
If a telehealth client just told you they want to triple spend for Black Friday, the next two weeks are when tracking problems get expensive. A broken dedupe that cost you a few percent of accuracy at normal spend now costs real money every day. An event that Meta quietly blocks on Wednesday takes your best ad sets back into learning right before the sale.
Here's what I'd check in every telehealth account before the budget goes up. Most of it takes a few minutes. None of it requires touching live campaigns.
1. Find out how Meta has categorized the pixel
Open Events Manager, select the pixel, and go to Settings. Look at the category and at Manage data source categories. If the data source is labeled health and wellness, Meta is already limiting what it accepts from that domain, and the limits depend on the restriction level.
Many telehealth brands sit in a partial restriction tier, where lower-funnel standard events like Purchase can't be used for optimization. If you didn't know which tier your client is in, that's the first thing to find out, because it decides what you can optimize to for the sale.
2. Check which events are blocked right now
In the same settings area, review event blocking. Write down every blocked event. Then compare that list with the optimization events in the campaigns you're about to scale. You don't want to discover during the sale that your Black Friday campaign is optimizing to an event Meta stopped using.
3. Look for tags on intake, quiz and checkout pages
This is the one that matters most outside the ad account. Load the client's intake flow in a private window with the browser's network tab open. Look for requests to Meta, TikTok, Snap, Google, Microsoft, Pinterest, Reddit or X while you move through the quiz and checkout.
The risk is a tracking request that pairs an identifier with a page or answer that implies a health condition. A URL like /weight-loss/intake/step-3 going to an ad platform says more than anyone intended. That combination is the theory behind the FTC's July 2026 complaint against Hims & Hers. The complaint's allegations haven't been proven, and Hims disputes them. But you don't need to wait for a ruling to see the risk.
4. Confirm pixel and server events are deduplicated
If the account sends the same event from the browser and from the server, both copies need the same event_id. Use Test Events and run a test conversion. You should see the browser and server versions merge. If they don't, reporting will overcount, and the algorithm is learning from inflated numbers.
5. Make sure every ad set can get out of learning
Meta's guidance is that an ad set needs roughly 50 optimization events in a week to leave the learning phase. When budgets jump, the temptation is to launch a pile of new ad sets for the sale. In a restricted telehealth account, where your optimization event is already further from revenue, splitting spend across too many ad sets means none of them stabilize before the sale ends.
Fewer ad sets, more budget each. Make significant edits before the sale, not during it.
6. Know which URL data Meta is stripping
Under core setup restrictions, Meta doesn't accept custom parameters or URL parts beyond the domain, including UTMs. If your reporting depends on URL parameters reaching Meta, it's already incomplete. Make sure the client's own analytics captures campaign source on their side, so your Black Friday read isn't based on half the data.
7. Choose the optimization event on purpose
If Purchase is blocked, don't let the account drift to whatever event happens to be available. Decide deliberately. Usually the best choice is the event that's closest to revenue and least revealing about a person's health, such as a completed checkout step or a qualified lead.
I'd skip the advice to rename a blocked Purchase event into something cryptic. I wrote about why separately.
8. Re-read the creative claims, especially for GLP-1
Tracking isn't the only thing that changes when spend goes up. Your ads get more eyes, including regulators'. The FDA sent 30 warning letters to telehealth companies about compounded GLP-1 marketing on March 3, 2026, and another round followed in June. Language reported as flagged in those letters includes calling compounded products "generic," "the same as" the brand-name drug, "clinically proven," or made in an "FDA-approved facility."
Run every Black Friday ad and landing page against that list before launch.
9. Compare the privacy policy with the actual data flows
Pull up the client's privacy policy and the homepage claims ("private," "discreet," "your data stays with your provider"). Then compare them with what you found in step 3. If the site says one thing and the network tab shows another, that gap is the issue regulators have focused on. The Hims complaint is about promises not matching data flows, not about pixels alone.
This isn't a conversation an agency should have alone. Flag it to the client and suggest they involve counsel.
10. Plan what happens if an event gets blocked mid-sale
Write down the fallback now. If your optimization event is blocked or degrades on day two of the sale, which event takes over? Who makes the call, and how fast? A two-line plan agreed in advance beats a scramble at 11 p.m. on Thanksgiving.
The short version
Before raising budgets: know the pixel's category, know what's blocked, check the intake flow for tags, test dedupe, consolidate ad sets, check the creative claims, and make sure the privacy policy matches reality.
If you want a faster start on step 3, Kalman's free scanner shows which ad-platform trackers load on a public site and flags pages that look risky, in about a minute. It's a good first pass before you open the network tab yourself.
This post is general information, not legal advice.