Free tiers, reverse trials, a-ha moments, and the hidden COGS of generosity in AI products.
Slack's 2,000 messages. Dropbox's cross-device sync. How these companies found their conversion-predicting product moments — and the cohort analysis methodology to find yours.
Only 6% of SaaS uses reverse trials. Start everyone on premium, downgrade after 14 days. The endowment effect does the selling for you. Here's why and how to implement it.
Low interest rates made generous free tiers cheap. AI inference made them expensive. Most companies haven't updated their math.
A reverse trial starts every new user on the premium/paid tier for a limited period (typically 14 days), then downgrades them to the free plan. The endowment effect — people value what they already have more than what they might get — makes this significantly more effective than traditional free-to-paid trials.
Analyze cohort data to find the behavioral threshold that correlates with conversion and retention. Slack found theirs at 2,000 messages per team; Dropbox found theirs when users synced a file across devices. The aha moment is the action that proves the product's value to the user.
It depends on inference costs. If your free-tier users generate meaningful AI compute, the COGS can be 10-50x what traditional SaaS free tiers cost. Many AI companies are tightening free tiers — adding rate limits, lower-quality models for free users, or replacing free plans with low-cost starter plans.
Freemium offers a permanently free tier with limited features or usage. Free trials give full access for a limited time. Freemium builds a large funnel and relies on self-serve upgrade; free trials create urgency but require credit card upfront or aggressive conversion flows.