Why decide in advance
Once a campaign is spending, your brain looks for reasons to wait: "a couple more hours", "there was an order yesterday". A rule written before the launch removes the emotion: the number is hit — the action is taken.
Rule 1. Set the test budget in payouts, not days
A handy starting point is a test budget per creative of 2–3 offer payouts. A $15 payout means a test of up to $30–45. If that money brings no leads and the click isn't cheaper than average, the creative goes off.
Rule 2. Switch off on a signal, not a mood
| Signal | What it means | Action |
|---|---|---|
| 2–3 payouts spent, 0 leads | The creative or angle doesn't work | Switch the creative off |
| Cheap clicks, low CR | Creative ↔ landing page gap | Change the page, not the budget |
| Leads come in, approval drops | Traffic quality | Narrow the audience or source |
| EPC consistently below CPC | The setup loses money | Switch the setup off |
Rule 3. Scale in steps
A working setup is easy to break with a sharp budget jump: the auction and the platform's algorithm react unpredictably. It's safer to add budget in steps of roughly 20–30% and check that EPC holds.
Test plan template
Hypothesis: a "before/after" angle will lift landing CR above 4%.
Budget: $45 per creative, 3 creatives.
Stop: $45 with no leads, or EPC < CPC a day after 20 leads.
Success: EPC > CPC with approval above 50% → budget +25%.
During onboarding you send a plan like this before every launch, and we review it together with the result. In a few weeks stop rules become a habit — and that's a buyer's most valuable skill.