Case study · Messaging campaign
588 conversations for $53
Creative and audience iterated weekly against cost per conversation, with the losing variants paused rather than left to average out the account.
- Conversations started
- 588
- Cost per conversation
- $0.09
- Amount spent
- $53.55
Summary
Creative and audience iterated weekly against cost per conversation, with the losing variants paused rather than left to average out the account.
Decisions that made the difference
- Reviewed creative and audience every week against one number: cost per conversation
- Paused losing variants instead of letting them dilute the account average
- Read the dashboard, not the promotional badge, when the two disagreed
How to read this result
Platform-reported. The poster’s badge reads $53.35; its own dashboard reads $53.55, and the dashboard wins. Figures are reported exactly as the platform shows them. The client name is withheld until written permission and a dated account export are in place.
More results
Other case studies
798 website purchases on a $1,052 budget
- Website purchases
- 798
- Cost per purchase
- $1.32
- Amount spent
- $1,052.28
Platform-reported in Meta Ads Manager. No attribution window stated on the export.
Read the methodSales campaign307 purchases at $1.44 each
- Website purchases
- 307
- Cost per purchase
- $1.44
- Amount spent
- $442.26
Platform-reported. The poster’s badge reads $462.26; its own dashboard reads $442.26, and the dashboard wins.
Read the methodSales test101 purchases under a dollar each
- Website purchases
- 101
- Cost per purchase
- $0.93
- Amount spent
- $93.54
Platform-reported. A test-scale budget — a read on unit economics, not on capacity.
Read the methodWant results you can audit?
Give us read access to your ad account. We will tell you where spend is leaking, what is mistracked, and what we would change first.
