Finding USD 100K in Marketing Waste with Incrementality Testing

Marketing teams often optimize to last-click metrics. That tells you what happened, not what would have happened anyway. Incrementality testing closes that gap — and at Careem it pointed to real savings.

The problem

Cross-channel spend was growing across TikTok, Meta, and Google. Each platform reported conversions. The question was not whether ads drove some activity, but how much incremental lift each channel actually contributed.

What we ran

Incrementality A/B tests designed to isolate true marketing lift — holding out control groups where appropriate and measuring incremental conversions rather than attributed ones.

The finding

The work identified a potential annualized burn reduction of USD 100,000 — spend that was not delivering incremental return at the margin.

Why it matters

Incrementality is unglamorous compared to creative optimization, but it is where analytics pays for itself fastest. You do not need a bigger budget; you need to know which parts of the budget are doing work.