How the plan is calculated
Results needed = number of ads × results per ad
Test budget = results needed × cost per result
Days needed = test budget ÷ daily budget (but never fewer than your minimum days)
The planner assumes spend is shared fairly evenly between ads. In practice Meta shifts budget towards early winners, so some ads get far less spend. If you need an even comparison, test ads in separate ad sets with equal budgets, or use Meta’s A/B test feature.
How many results per ad is enough?
There is no magic number, because it depends on how different the ads really are. As a rough guide:
- Under 10 results per ad: mostly noise. Treat any “winner” as a hunch.
- 20–30 per ad: enough to spot large differences (one ad converting at half the cost of another).
- 50+ per ad: needed to separate ads that perform within 20–30% of each other.
Meta’s own learning phase guidance (roughly 50 optimisation events per ad set per week) is a useful reference point for how much data the system wants.
Make tests worth the money
- Test meaningfully different ideas: a new hook, offer, format or audience problem. Small tweaks need far more data to measure.
- Change one major thing at a time so you know what caused the difference.
- Decide the judging metric and the end date before you start, and stick to them.
- When budgets are small, judge on an earlier signal with more volume, such as click-through rate or cost per landing page view, then confirm the winner on conversions.
Once a winner is running, watch for fatigue with the Ad Frequency & Fatigue Checker.