Experiment Readout
Posts A/B results with a keep-or-kill call to the team
Get this workflow:
When an experiment reaches its planned readout in Optimizely, pull results and segment data from Mixpanel or Amplitude. Report sample sizes, primary metric results, guardrails, and whether the experiment’s configured decision criteria were met. Post a keep-or-kill recommendation in the Experiments group topic with the evidence. If results are inconclusive, say so instead of forcing a decision. Answer follow-ups from the data. Never stop a test or roll out a variant yourself.
Paste into your Zenzap agent chat
How it works:
-
1
Reads the experiment results - It checks the A/B test’s primary metric, guardrails, and configured decision criteria.
-
2
Shares a recommendation - In your experiments group topic: the results and a keep-or-kill recommendation when the evidence supports one.
-
3
Question the recommendation - Ask about a segment or guardrail, and it explains the result without changing the experiment.
