Allocation pattern · B2C SaaS

Wellness App: Installs Up 40%, Paid Subscribers Flat

A seed-stage consumer subscription app came in last quarter. Mental wellness category, $9.99 monthly or $59.99 annual subscription. $2M ARR. The founder’s deck showed cost-per-install down 28 percent quarter over quarter, install volume up 40 percent. Paid subscribers up 6 percent. The math did not add up.

The agency was reporting on the install funnel. Free-to-paid conversion sat at 4 percent against an 11 percent target. 60-day churn on paying subscribers was running 78 percent on TikTok and 52 percent on Meta. The dashboard looked clean. The revenue did not.

The diagnosis (3 sentences)

The platform algorithms were optimizing for the event the founder had told them to find. The event was “install,” which is abundant, cheap, and uncorrelated with paying-subscriber behavior. The audience getting cheaper to acquire was an audience that installs apps and uninstalls them.

The allocation move

  • Switch the optimization event on Meta and TikTok from “install” to “7-day paid trial start.” Server-side event tracking via the subscription infrastructure (RevenueCat or similar) is the technical prerequisite. 2 to 4 weeks of engineering for the integration.
  • Expect cost-per-event to rise sharply at the new optimization layer. Install cost was around $2.80 on Meta; cost-per-trial-start jumped to around $14. The math is fine because paying-subscriber CAC drops from ~$90 blended to ~$40 within 60 days.
  • Reallocate 30 percent of TikTok budget to Meta where the 60-day cohort retention was twice as strong. TikTok stays in the mix at lower share for creative angle development at the top of funnel, not for direct paid-subscriber acquisition.
  • Refresh the creative angle library against a buyer-state matrix. Three problem-aware angles (“can’t sleep”) had saturated. Add four new angles addressing solution-aware and brand-aware user states (identity, ritual, status).
  • Rebuild the LTV projection using actual cohort retention curves, not the optimistic curve. The blended 12-month retention was 38 percent against a planning assumption of 60 percent. Recalculate CAC payback against the real number, not the planning one.

When this applies. When it does not.

Applies: B2C SaaS consumer subscription apps in wellness, mental health, sleep, fitness, meditation, or similar categories. Paid is optimizing for install or free signup. Free-to-paid conversion below 6 percent. Cost-per-install looks healthy on the dashboard while paid subscriber growth is flat.

Does not apply: Apps with ad-based monetization (where install volume is the revenue driver). Apps already above 10 percent free-to-paid conversion (the optimization event is fine; the gap is scale). Pre-PMF apps with sub-100 paying subscribers (too early to optimize the funnel; the audit will tell you to keep validating before deploying paid).

A founder line, anonymized

“I had been paying Meta to find people who would install my app. I needed Meta to find people who would pay me $60 a year. The audit explained why those are two different audiences and how to switch the algorithm to the second one. Free-to-paid went from 4 percent to 11 percent in 90 days. Paid spend stayed flat.”

For the full 5-check B2C SaaS paid diagnostic, see B2C SaaS Paid Strategy: Why Cost-Per-Install Lies. The optimization-event fix is the highest-leverage change in subscription-app paid that I see in audits.

Book the 10-Day Allocation Audit.