How a Series A AI calendar app got 9,400 trial starts in 8 weeks

The company and program

The approach

• Hypothesis: A creator who connects their real calendar and lets the AI plan their week on camera drives more paid conversions than a feature walkthrough.
• Comparison: 5 "my real week" videos against 4 feature-tour videos in the first wave, each with its own tracked link. On paid, creator-handle ads against the app's earlier Meta brand ads at $118 per paying customer.
• Decision rule: Paid conversion takes 14 days, so ad decisions used a leading signal. After 7 days of organic data, ads went behind videos whose trials connected a calendar in the first session at 70% or more. Each ad concept got $500–$1,000 or 7 days. Rebooking waited for day-14 trial-to-paid, with a bar of 25%.

What we changed

1. Weeks 1–2: shortlist and real accounts. We reviewed 110 channels, contacted 28 and signed 9 for the first wave: workplace productivity tutorial channels, note-system teachers and step-by-step software tutorial channels. Each creator got an account a week before filming and connected their real work calendar.
2. Week 3: 9 videos go live. Five "my real week" videos and four feature tours, each with its own link.
3. Weeks 4–5: read and amplify. "My real week" trials connected a calendar in the first session 78% of the time, against 66% for feature tours. The top 3 videos ran as creator-handle ads, with 90 days of paid usage in each base fee.
4. Weeks 6–8: second wave. Four "my real week" creators posted a "30 days later" follow-up, and 2 new creators joined, for 11 creators in total.
5. Weeks 9–10: close the cohorts. The last trials reached day 14, and we read paid conversion by creator before rebooking for Q4.

The results

What didn't work

• Feature tours. The 4 feature-tour videos converted trials to paid at 17%, under the 25% bar. None was rebooked.
• Broad reach without a workday audience. Our largest booking, a self-improvement channel with over 1M subscribers, brought 1,050 trials but only 16% paid. Its viewers were students more than people with full calendars.
• Android lagged. Calendar sync on Android was slower, and Android trials converted at 15%, against 29% on iOS and Mac. We asked creators to film on iPhone and Mac and pointed Android viewers to the waitlist until sync improved.

What we learned

• Have every creator use their real calendar, not a demo account.
• Make ad decisions on a fast leading signal, and rebooking decisions on paid conversion.
• Book "30 days later" follow-ups with the creators who converted best.

Frequently asked questions

How much did 9,400 trials from YouTube creators cost?

Each trial cost $12.13 and each paying customer $44.92, against $118 on the app's earlier Meta brand ads. Trials from creator links cost $53.35 per paying customer; trials from creator-handle ads cost $25.77.

How long before creator videos turn into paid subscribers?

Videos went live in week 3. The leading signal, calendar connected in the first session, read at 7 days. Paid conversion waits for the 14-day trial to end, so the full 27% read closed in weeks 9–10.

Would this work for a Series B productivity app, or an Android-first one?

A Series B app with a 14-day trial and a workday audience, yes: the real-calendar format and the 70% leading signal carry over. Android-first, not yet. Android trials here converted at 15% against 29% on iOS and Mac, so fix sync first.

Related case studies and guides

Same industry, other angles:

• Paid ads: Paid subscriptions up 3.2x in 4 months at a flat $52 CAC
• Product growth: Workspaces inviting a second member up from 19% to 37%
• Full growth system: A Series A AI meeting assistant grew paid seats from 9,000 to 34,000 in 8 months

• Creator marketing for productivity apps
• YouTube tutorial campaigns for productivity apps
• Creator lists by product category