Tailor AI · Help · Concept
‘Too early’, what it really means
From the Tailor AI team · Reviewed
Answer
‘Too early’ means at least one requirement is unmet: 200 total impressions across the compared arms, 1 conversion in each arm, 10 conversions in at least one arm, and 5 full days spent testing.
All four requirements must pass before a test clears ‘Too early’. Time paused or after rollout does not count toward the five days. An unknown runtime also stays ‘Too early’. After those gates pass, the confidence score determines the label: below 75% is Low Confidence, 75% to below 90% is Medium Confidence, and 90% or above is High Confidence. An account may require a longer runtime before a winner or loser is called. For pipeline or revenue goals, allow for conversion lag.
What I'd do next
- If volume is low, collapse to A/B and simplify segments.
- If the goal is extremely downstream, add a reliable leading goal as secondary.
Caveats
If your goal tracking is broken, it can look like ‘too early’ forever.
Related questions
Trusted by growth and marketing teams at leading B2C and PLG companies
See Tailor on your site
Preview how Tailor adapts your pages to traffic intent.
Looking for setup guides instead? Browse the documentation.