How Zapier achieved 190% ROAS uplift with pLTV optimization

The Challenge

Zapier was acquiring users at scale but struggling to distinguish high-value subscribers from one-time users at the point of acquisition. Their ad platforms were optimizing for trial signups, not long-term revenue.

The Solution

Churney connected to Zapier's data warehouse and trained causal ML models on their historical subscription data. The models predicted each new user's expected lifetime value based on early engagement signals.

The Results

With pLTV scores flowing to Meta and Google in real time, Zapier's ad platforms could bid more aggressively for users predicted to have high lifetime value. The result: 190% improvement in ROAS, measured through controlled incrementality tests over a 90-day period.