Sub Club by RevenueCat · 2025-12-24
Creative Misfires, False Positives, and Meta's Auction Flaws — Alper Taner, Stealth-Mode App Studio
Alper Taner, head of performance marketing at a stealth app studio, on treating best practices as inputs to test, treating weak metrics as facts until tested, interpreting creative results one variable at a time, setting creative spend thresholds from cumulative CPX curves, rebalancing ad sets dominated by one ad or country, a capped 'inflated budget' Meta setup, and remapping a trial event from Purchase to trial on Meta, which cut cost per trial ~35%.
Play the episode11 tactics
Numbers mentioned
| Subject | Metric | Value | Quote |
|---|---|---|---|
| personAlper Taner, annual UA budgets managed | ad spend | eight figures approximate | “Alper has been in mobile growth for over a decade, leading UA and MarTech efforts across a wide range of verticals and managing eight-figure annual budgets.” |
| benchmarkHealthy install-to-trial-start rate for subscription apps | conversion rate install -> trial start; contrasted with a 5% account | 15–20% approximate | “And usual healthier benchmark is more like 15% to 20% and such.” |
| benchmarkMeta's early delivery decision window for a new ad | count Meta makes delivery decisions in roughly the first 5,000-10,000 impressions; ads can be starved too early | 5,000–10K impressions approximate | “because Meta does some decision making around first 5,000 to 10,000 impressions” |
| aggregateAccounts Alper has seen letting ad sets run with one dominant ad | share Share of ad-set spend going to one dominant creative or country, the trigger to rebalance | 80–90% approximate | “When we see a 80%, 90% of a dominant asset, let it be a creative, let it be a country and such, you're not fulfilling the potential of the other variables.” |
| Large app account with no spend guardrails | ad spend Spend on a single creative running at a $100+ CPI before anyone stopped it; should have been killed at $2,000 max | USD 90K–100K approximate | “I've seen as crazy as someone spend 100K on a creative with $100+ and did not stop it before. I think it was like 90K something didn't stop.” |
| Large app account with no spend guardrails | cpi CPI of the runaway creative ($100+) | USD 100 approximate | “With $100 cost per install?” |
| Large app account with no spend guardrails | cost Spend level by which the bad creative should have been killed | USD 2,000 estimate | “it was high enough that you should have killed it after $2,000 max and not wait for 100,000” |
| Account where Alper scaled a winning Meta creative | growth rate Campaign budget growth within two weeks with same or falling cost per purchase | +5 x approximate | “And we were able to, I think 5X the campaign budget within two weeks with the same and even decreasing cost per purchase.” |
| benchmarkStarting bid/cost cap relative to current cost per purchase in a capped inflated-budget setup | share Start caps 10-20% below current cost per purchase (e.g. ~$45 when CPP is $50), then raise slowly | 10–20% approximate | “If your cost per purchase is 50, you want to start, I don't know, around 45 or 10%, 10% to 20% below and you might not get delivery, then you play it slowly.” |
| benchmarkAccount size where the capped inflated-budget tactic makes sense | ad spend Large accounts already spending ~50K/day that want to reach 100K/day | USD 50K/day approximate | “this works well at large accounts with large learnings that they're spending 50K a day anyway type of a thing and they want to see if they can spend more than 50K a day” |
| benchmarkPerformance gap between placements/demographics that justifies testing Meta bid modifiers / value rules | share | 20–30% approximate | “If you have performance difference above 20%, 30% among the, let's say the placements or demographics and such, it is worth testing it.” |
| benchmarkMonthly UA spend above which incrementality measurement (MMM, geo-lift, holdouts) becomes important | ad spend | USD 500K/month approximate | “And once you get to certain level, I would say 500K+, then these are becoming a lot more important and prominent in decisions” |
| Subscription app with high CAC that optimized Meta for purchase | cpa Cost per trial after mapping the trial event to a trial event on Meta instead of Purchase; same creatives and campaigns | +-35% approximate | “all we did was mapping their trial event to trial and their cost per purchase/the actual cost per trial went down by, I think, around 35%, only with that change.” |
| benchmarkTime to let a newly created Meta event build history before split-testing it | time to milestone | 1–2 weeks estimate | “so I would at least wait a week or two so that the events manager will build some history” |