# My experiment ran for weeks and still shows no winner, what now?

Asked by sam.qa on 2026-11-04. Tags: ab-testing, experiments, statistics.

trying to figure out if something is wrong before i adress it with my lead.
here is what i checked:

- both variants are well over 1000 visitors (about 4k each)
- both are over 25 conversions
- it has been running almost three weeks

so the guardrails are clearly met, but neither variant crosses the recommended
winner bar. probability to be best is sitting around 60/40 and just will not
move. is the experiment broken or is this a real outcome?

## 0 answers