What should my metrics look like on day 1, day 3, and day 7 of a new test?
The #1 gap in the course — operators kill campaigns at day 1 before learning phase has any data
This is the question the community asks most and gets wrong most expensively. Here are realistic benchmark ranges for a new product test:
- Day 1: Expect underperformance. CPM is higher than it will be (learning phase). CTR may be lower. CPA may be 2–3× your target. This is normal. Do not make kill decisions on day 1 data. Only kill if you have zero ATCs after $50+ spend at a correct pixel setup — that indicates a fundamental creative or offer problem.
- Day 3: Learning phase is partially complete. You should start seeing ATC events. If ATC rate is below 2% after $100 spend across multiple creatives, check your product page. If ATC is above 3% but no purchases, check checkout friction (shipping cost, payment options, trust signals).
- Day 7: Post-learning data. This is where real decisions get made. If ROAS is below break-even after $300–$400 total spend, you have a real signal. If ROAS is above break-even, scale cautiously. If ROAS is borderline, evaluate: which angle got the most ATCs? Iterate on that angle with new creative variations before killing.
The single most expensive mistake in ecom: killing a campaign between day 1 and day 3 based on no purchases. You're destroying the most expensive data you'll ever collect — Meta's learning phase data costs real money and cannot be refunded.
Tactical FixSet a calendar reminder for day 7 when you launch a new test. Your kill/scale decision window is day 7, not day 1. Check daily for catastrophic failures (zero ATCs after $50, broken pixel, checkout not working) but make no structural changes until day 7 data is in.
See this in practice: Read Your Data →
Related Content
FAQs
What does a healthy account look like in terms of metrics across all active campaigns? →My ad metrics look fine but I'm getting zero add-to-carts. Where is the traffic going? →When Meta's data says one thing and Shopify says another — which do I trust? →What does it mean when my CTR is good but my CPC is still high? →
More on Reading Ad Data
- My frequency is going above 3. Should I be worried?
- What's a healthy hook rate and why does it matter more than I thought?
- When should I trust my data and when is it just statistical noise?
- How do I know if my CPA is acceptable when I don't have a ROAS target yet?
- My ROAS looks great but I'm not profitable. What's happening?
- I have purchases but my ROAS looks different in Meta vs Shopify. Which do I use to make decisions?
