Practitioner pattern · Alper Taner
Plot each ad's total spend against its cost per result to see when to stop
Bad ads are often clear by ~$500; one account burned ~$90K on a $100+ CPI ad
Workedpaid-socialexperimentation
What they did
There is no universal per-creative test budget. For your account, chart cumulative spend against cumulative CPX (CPI, cost per trial or cost per subscription) across past creative tests and find the spend where the metric stabilizes; use that as the decision threshold, not statistical significance and not $20-30. Add automated guardrails/rules so large accounts don't keep spending on a clear loser.
What happened
Stops spending 10-20K on creatives that were clearly bad after ~$500.
In their words
Sub Club by RevenueCat · Creative Misfires, False Positives, and Meta's Auction Flaws — Alper Taner, Stealth-Mode App Studio“But what I would suggest is to look at your cumulative spend and cumulative CPX metrics, whether you're testing on a CPI or a cost per trial or just cost per subscription and such, and see when they stabilize.”
“I'm like, "Oh, really? I could tell it after 500 bucks, look at the data. It was already bad. It was already screaming. Why did you spend so much?"”
Related topics
Get tactics like this every Monday
New tactics from the week's founder interviews, each linked to where it was said.
More from this episode
- When an ad wins, change one thing at a time to learn why before making moreExpert says it creates more winners; seen a concept's third variant win after two flopped
- Track cost per new ad that beats your current ads, not how many you testExpert says fewer, better-reasoned tests beat 500-a-month volume; no numbers
- When one ad eats 90% of an ad set's budget, give the other ads their own ad setOther ads often 'bloom' once the dominant ad is removed
- Keep the US out of campaigns with smaller countries; group countries by cost or value
- When a new ad keeps overspending its budget at the same cost, raise it fast5x campaign budget in two weeks with falling cost per purchase
- Set a $1M daily budget on Meta, but cap total spend and bids to control itMore high-quality traffic at stable cost for some accounts; others got no or runaway delivery
- Bid differently by placement or age group when their results differ by 20%+
- Tell Meta a trial start is a trial, not a purchaseCost per trial down ~35% with the same ads and campaigns
- Use MMP data to compare ads, but switch channels off to see what they really add
- Check which competitor ads have the most versions and longest runs in Meta's ad library