Two numbers decide whether a test can teach you anything: how much each slice needs to spend, and when a slice has proven itself dead. Get either wrong and you are not testing — you are gambling in smaller increments.
Per-slice budget = 3 × payout · Total budget = slices × per-slice budget
The 3× payout rule buys enough clicks for several expected conversions per slice. At a 2% conversion rate and a $60 payout, $180 buys roughly 225 clicks at a $0.80 CPC — about 4–5 expected conversions, enough for a real signal. Cut slices, never slice budgets, when money is tight.
If a slice truly converts at your expected rate, the chance of seeing zero conversions after n clicks is (1 − CVR)ⁿ. The kill threshold above is the click count where that probability drops under 10% — at 2% expected CVR, that is 114 clicks. A slice that burns past its kill threshold with no conversion is statistically unlikely to be a winner, no matter how good the creative looks. Kill it and reallocate.
Already know your numbers and just need the click-price ceiling? Use the companion break-even CPC calculator.
Plan for at least 3× the offer payout per targeting slice — one GEO, placement, or creative combination. For a $60 payout that is $180 per slice, and a realistic first test covers 3–5 slices, so $600–$1,200 total. Testing below 3× payout per slice means random variance, not data, decides whether you keep or kill a slice.
A targeting slice is one isolated combination you judge on its own numbers: a single GEO, placement or zone, and creative pairing. Budgets and kill decisions belong to slices, not to whole campaigns, because a losing GEO can hide a winning placement inside averaged campaign-level data.
Kill a slice when it spends your per-slice test budget without reaching the conversions your expected conversion rate predicts. Mathematically, after n clicks with zero conversions, the probability of that happening by chance is (1 − CVR)^n. At a 2% expected CVR, 114 clicks with zero conversions means there is under a 10% chance the slice truly converts at 2% — that is a data-backed kill point.
Break-even CPA equals your payout times your approval rate — the cost per action where profit is exactly zero. Max viable CPA subtracts your target profit margin: break-even CPA × (1 − margin). A $60 payout at 100% approval with a 30% margin target gives a $42 max viable CPA. Any acquisition cost above $42 spends your planned profit.
Yes, but only if each one gets its own full slice budget. Three slices at $180 each produce decisions; six slices at $90 each produce noise. If total budget is limited, cut the number of slices before you cut the per-slice budget — fewer slices with clean data beat many slices with ambiguous results.
Yes, completely free with no signup, and it runs entirely in your browser so no campaign data leaves your device. It is built by TrafficBlueprint, a campaign planning tool that scores traffic sources against your offer economics and flags compliance risk before launch.
TrafficBlueprint scores 22 traffic sources against your payout, margins, and vertical restrictions — so the test you fund has a chance to pass.
Plan your campaign free →