Experimentation
Compare two groups of URLs to see whether a change had a measurable effect.
Property → Forecasting → Experimentation1 min read
How to run it
- Create an experiment by uploading a CSV (
url,group) or auto-splitting by URL path or a random A/B split. - Optionally set a launch date; the groups are saved.
What you get
- Each group’s click growth against the prior period, with the winner highlighted.
- A statistical significance test on the difference, a confidence interval, and a low-sample power warning.
- When a launch date is set, a chart of the cumulative difference between the test and baseline groups since launch.
Results are observational (Search Console data is sampled and the first group is the baseline), so read them as strong evidence rather than a controlled trial.
Part of Forecasting & experiments. Related: SEO Traffic Forecast.