12-Month Forecast: When Does a Freemium-Only Launch Need to Add a Trial?
Objective: Given a 12-month signup and conversion trajectory for a freemium-only launch, forecast when paid-customer growth plateaus, and calculate whether layering an opt-out trial on top (a hybrid model) meaningfully changes the 12-month outcome.
Sea Limited is piloting a freemium seller-analytics tool inside Shopee's merchant app. Six months of real usage data shows freemium signups growing but paid conversion flattening. Leadership wants a 12-month forecast before deciding whether to add a premium trial on top.
Project month 7-12 paying customers under 'freemium stays freemium-only' versus 'freemium adds a hybrid opt-out trial in month 7,' using the lesson's benchmark rates, and recommend a path.
Before you start
What you'll need
Free path (everything below is enough to finish)
A month-by-month cohort table is a spreadsheet exercise, no forecasting software required
Paid upgrades (optional, faster/deeper)
Cohort-based conversion analysis over time is Amplitude's core use case once you have real usage events, not a synthetic export
No access? Export raw signup/conversion event timestamps from any analytics tool and build the cohort table in Google Sheets manually
The process
2 steps
Step 01 of 02
Freemium-to-paid conversion in this lesson's benchmark sits at 3-5%, and it compounds slowly because it depends on users hitting a usage ceiling over time, not a fixed deadline. Month-over-month paid growth from a flat signup base plateaus once the addressable pool of 'about to hit the ceiling' users is exhausted.
Months 1-6 show 4,000 new freemium signups/month and a steady 3.5% eventual conversion rate, but months 5 and 6 both added only ~40 net new paying customers versus ~140 in month 2. What's happening, and what does months 7-12 look like if nothing changes?
Procedure
- Extend the 6-month actuals at the same 4,000 signups/month and 3.5% eventual conversion rate for months 7-12.
- Note that new paying customers per month depend on how many freemium users are newly hitting the usage ceiling, not the whole free base, so growth flattens once early cohorts have already converted or plateaued as non-converters.
- Forecast: months 7-12 continue near the ~40-50/month plateau seen in months 5-6, absent any change to the funnel.
- Total 12-month paying customers under freemium-only: roughly 700-750.
Freemium-only forecast, months 1-12 Month 1: 4,000 signups, 140 new paying (early high-intent cohort converts fast) Month 2: 4,000 signups, 140 new paying ... Month 5: 4,000 signups, 42 new paying (ceiling pool shrinking) Month 6: 4,000 signups, 39 new paying Month 7-12 (forecast): ~40/month flat -> ~240 more 12-month cumulative paying: ~730
Healthy
Forecast is built from the observed plateau in months 5-6, not extrapolated from the strong early months.
Unhealthy
Assuming months 7-12 will keep adding 140 paying customers/month because that's what months 1-2 did, ignoring the visible plateau.
What this means
A freemium-only model's paid growth is bounded by how fast users hit the usage ceiling, not by signup volume. Once the early high-intent cohort converts, growth flattens even with steady signups.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| New paying customers per month dropped from 140 to ~40 by month 5, with signups holding flat | Flag the plateau to leadership before month 7 planning, don't wait for a full quarter of flat growth to notice | 5 min |
Step 02 of 02
The lesson describes the hybrid model (Canva, Loom, Webflow): a freemium base handles distribution, and a time-boxed premium trial layered on top handles conversion urgency, converting users who already built a habit on the free tier.
If month 7 adds a 14-day premium trial for freemium users who've hit the usage ceiling (targeting the plateaued ~3,650 free users from the first 6 months, not new signups), and that trial converts at the opt-out benchmark of 49.9% among the ~15% who actually start a trial, what does that add to the 12-month total?
Procedure
- Identify the plateaued pool: ~3,650 free users from months 1-6 who never converted under freemium alone.
- Apply a 15% trial-start rate (users who've already built a habit are the most likely to try a premium trial) = ~548 trial starts.
- Apply the 49.9% opt-out conversion benchmark to those trial starts = ~273 additional paying customers, one-time, in month 7.
- Add this to the freemium-only forecast: ~730 (freemium-only) + ~273 (hybrid trial layer) = ~1,003 paying customers by month 12.
- Note the caveat: this is a one-time unlock of the existing plateaued pool, not a new steady-state rate, month 8-12 still needs its own hybrid cohort math for a real model.
Hybrid transition, month 7 one-time cohort Plateaued free base: 3,650 users Trial starts (15%): 548 Converted at 49.9%: ~273 new paying customers 12-month total: 730 (freemium-only baseline) + 273 (hybrid unlock) = ~1,003
Healthy
The hybrid layer is modeled as a one-time unlock of the existing stuck cohort, with an explicit caveat that it isn't a repeatable monthly rate.
Unhealthy
Treating the +273 hybrid bump as a new permanent monthly run-rate and extrapolating it forward without re-deriving month 8-12 separately.
What this means
A hybrid model's biggest lift often comes from finally converting users who already built a habit but had no urgency mechanism, exactly the population freemium alone structurally can't monetize.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| Freemium-only forecast plateaus at ~730 paying customers by month 12 | Recommend layering an opt-out trial for the plateaued cohort in month 7, expected one-time lift of ~273 paying customers | dev ticket |
Final deliverable
A 12-month forecast comparing freemium-only versus a month-7 hybrid trial layer, with the one-time-unlock caveat stated explicitly, and a recommendation.
See a reference example
Peloton digital-only tier, 12-month hybrid forecast (excerpt) Freemium-only baseline: ~610 paying by month 12 (plateau visible from month 4) Hybrid trial layer (month 7, targeting 5,200 plateaued free users): +389 one-time Revised 12-month total: ~999 Caveat: month 8-12 requires a fresh cohort model, this is not a new run-rate.
Success criteria
You're done when you can:
- Correctly identifies the freemium plateau from the observed month 5-6 data rather than extrapolating early months
- Models the hybrid trial layer as a one-time cohort unlock, not a permanent new monthly rate
- States the caveat about needing a fresh cohort model for months 8-12 explicitly in the deliverable