Paid Media · Experimentation
Series B DevTools SaaS
Cut CAC by 65% by rebuilding the creative testing engine and pruning underperforming audiences.
-65%
CAC
steady
Trial volume
2 quarters
Time to result
Problem
CAC had crept up 3 quarters in a row as the team scaled spend without a structured way to test creative or audiences. Every campaign was a one-off; nothing compounded.
Approach
- Rebuilt the account structure around a creative-testing framework with clear statistical thresholds.
- Introduced a weekly experiment cadence across hooks, formats, and landing pages.
- Shifted budget from broad prospecting to intent-rich retargeting and lookalikes informed by product usage data.
Execution
- Shipped 40+ creative variants across Meta and LinkedIn in the first quarter.
- Built a lightweight AI pipeline to generate and score ad copy variants before they reached spend.
- Instrumented attribution to tie ad-level spend to trial activation, not just clicks.
Outcome
CAC dropped 65% over two quarters while trial volume held steady, freeing budget to reinvest into the channels that were actually compounding.