DTC telehealth · Growth & CRO
Fix the leaks before you buy more traffic.
In 16 weeks, we audited a fast-growing GLP-1 telehealth brand end to end, rebuilt how it measured and priced its offer, put every ad through compliance review, and caught nine revenue and regulatory risks before they shipped. All of it happened during a live platform migration.
- Client
- DTC telehealth brand prescribing compounded GLP-1 medications. Name withheld.
- Engagement
- Fractional growth and CRO, June to October 2026
- Our role
- Embedded growth lead: audit, measurement, pricing, lifecycle, creative compliance, brand system
- Stack
- WooCommerce · ActiveCampaign · GA4 · PostHog · Hyros, migrating to Stripe · Customer.io
~6%of real revenue the analytics stack was capturing when we arrived
1 defectbehind churn, disputes, bad reviews, and payment-processor risk
500+ad assets reviewed for compliance, including 102 creator videos checked line by line
9revenue and regulatory risks caught before launch
The challenge
The brand scaled fast on paid and affiliate traffic, then hit a wall.
Subscribers were churning, disputes were climbing, and public reviews had turned. Leadership couldn’t tell which channels were working.
- Two stacks at onceCustomers were moving from WooCommerce and ActiveCampaign to a new telemedicine platform on Stripe and Customer.io. Every metric, price, and email had to be right in both places.
- A tightly regulated categoryCompounded GLP-1s allow no weight-loss promises, and every claim needs substantiation and legal review. Tactics that work elsewhere create real exposure here.
- A fractional seatWe were brought in to fix email and stand up a new lifecycle platform. The real job was finding where the business was losing the most value, and fixing that first.
What we did
Six workstreams, ordered by leverage.
Diagnose first, then measure, monetize, protect, retain, and systematize. All six ran in parallel.
- 01 · DiagnoseOne audit, one root cause.
- Merged three audits into one 25-page report with a prioritized action plan.
- Traced churn, disputes, collapsing reviews, and processor risk to one defect: no self-service way to cancel. One inexpensive fix addresses all four.
- Showed a sharp mid-year decline was three pressures landing at once, and would repeat with later cohorts unless retention was fixed.
- 02 · MeasureNumbers you can trust.
- Analytics captured about 6% of the revenue in the order ledger, so every channel metric was a small sample.
- Found why cross-domain purchases weren’t tracking: 5 purchase events logged against 149 paid checkouts.
- Caught an affiliate cookie that overwrote itself on every visit, and set one revenue source of truth for the migration.
- 03 · MonetizePricing and plan mix.
- Modeled lifetime value from 18 months of orders. Kept the monthly plan and specified an upgrade prompt worth about +$185 LTV per upgrade.
- Ran the go-live audit for a move to always-on pricing: 12 surfaces across two platforms, a launch runbook, and FTC-safe claims.
- Showed the post-change dip was timing, not price: cash per monthly order rose from $140 to $191.
- 04 · ProtectCompliance review on every ad.
- Ran the review gate with the media agency: 500+ assets across a dozen rounds, with written fixes on each.
- Checked all 102 creator videos line by line for price and disclaimer, and scored a 66-creator roster.
- Flagged three launch-blocking issues on the live funnel.
- 05 · RetainLifecycle and email.
- Drafted five full flows for the new platform, including win-back for a 215K-contact list.
- Audited past sends: 22% to 63% of clicks per email came from security scanners, not people.
- Rebuilt the next campaign around engaged segments and a 10% holdout for true lift, recovering about 400 real buyers the platform’s filters missed.
- 06 · SystematizeA brand system agencies can follow.
- Found fonts from another brand bleeding into the live landing page, and anchored type to production.
- Built a design system (149 component families, 132 tokens) and an 18-slide brand guideline with pricing and compliance rules built in.
What we caught before it cost money.
Each one was found in the data or on the live site, documented with evidence, and routed to an owner.
| Risk | What we found | Why it mattered |
|---|---|---|
| Renewals not billing | The new platform had never completed a subscription renewal | About 40 first renewals were due within three weeks. No renewal meant no refill for patients. |
| Coupon over-discount | A retired first-order coupon still worked against the new lower prices | On one product it would have pushed orders below zero |
| Fake scarcity | A countdown and “discounts left” widget that reset on every visit | A recognized FTC dark pattern that contradicted the brand’s transparent-pricing promise |
| Outcome claim at checkout | A dated, quantified weight-loss promise in the checkout headline | Off-limits in this category. A footnote doesn’t cure a headline. |
| Ad claims | Shipping and cancellation language operations couldn’t yet back up | Unsubstantiated claims in paid media, at scale |
| Ad whitelisting | 13 pages with clustered IDs, generic personas, and no overlap with real creators | Held until ownership was verified |
| Payment-processor risk | Dispute volume high enough to put the merchant account at risk | Losing the merchant account would stop all revenue |
| Untested annual plan | The highest-priced, highest-refund-risk plan had never processed an order | We wrote the full test checklist before any traffic reached it |
| Underpowered split test | At current volume, a homepage test needed about 10,000 sessions per arm | Reset the primary metric and run length so the result would mean something |
Growing fast and not sure what’s working?
We’ll find the leaks first. Your first conversation is with Dan.