The Challenge
TaskFlow (name changed) is a B2B project management SaaS with 10,000 monthly trial signups. Their trial-to-paid conversion rate had plateaued at 8%, and the product team had a backlog of 50+ ideas to improve it—but no way to test them.
Key Problems
- No experimentation infrastructure: Feature flags were managed through config files with no measurement
- HiPPO decisions: Product decisions were made by opinion, not data
- Unclear activation: No visibility into which features drove conversion
- Slow feedback loops: Changes shipped, but impact was unknown for months
The Solution
We implemented PostHog as their product analytics and experimentation platform, enabling data-driven product development.
Architecture Overview
- PostHog Cloud: Event tracking, funnels, cohorts, and session recordings
- Feature Flags: Percentage rollouts and user targeting
- Experimentation: A/B testing framework with statistical significance
- Activation Analysis: Identifying features that predict conversion
Implementation Process
Phase 1: Instrumentation (2 weeks)
We instrumented their React app with PostHog, tracking 35 key events across the trial journey. We implemented proper user identification and group analytics for workspace-level analysis.
Phase 2: Activation Analysis (1 week)
Using cohort analysis, we identified the "activation moment"—users who created a project AND invited a teammate within 3 days converted at 3x the rate of others.
Phase 3: Experimentation Program (ongoing)
We established an experimentation framework: hypothesis → feature flag → measure → decide. The product team now runs 3-4 experiments per sprint.
Winning Experiments
Over 6 months, 12 experiments ran. Here are the winners:
Experiment 1: Onboarding Checklist (+12% conversion)
Added a visible checklist guiding users to activation. Users who completed the checklist converted at 22% vs. 8% baseline.
Experiment 2: Team Invite Prompt (+8% conversion)
Moved the "invite teammates" prompt from settings to the main workspace. Invites increased 40%, driving activation.
Experiment 3: Trial Extension Offer (+6% conversion)
Offered a 7-day trial extension to users who hadn't activated by day 10. 30% took the extension, and 15% of those converted.
Experiment 4: Premium Feature Preview (+5% conversion)
Showed premium features (greyed out) during trial with "Upgrade to unlock" badges. Increased upgrade intent by making value visible.
Results
Compound effect of winning experiments:
- 31% higher trial conversion: From 8% to 10.5% trial-to-paid
- $1.2M additional ARR: 250 more paying customers per month at $400 ARPU
- Faster iteration: Product team now ships experiments weekly
- Data culture: "What does PostHog say?" is now a standard question
"Before PostHog, we shipped features and hoped they worked. Now we test everything. The 31% conversion lift is great, but the bigger win is that our entire team thinks in experiments."