How to use the SEO Experiment Planner.
Define the change, comparison group and observation window before editing. Search performance can change for reasons unrelated to the experiment.
Make the workflow fit your task.
Define the page change, comparison cohort and observation window before editing. Record baseline behavior and possible confounders, then choose the evidence that would support keeping, revising or reversing the change.
- What you provide
- Page cohort, hypothesis and available metrics.
- What you get
- Change plan with comparison method and measurement limits.
See the input and the result.
Illustrative input and output · a teaching example, not a live WebAct run
Example input
Plan an SEO test of clearer titles on comparable tool pages. No baseline data, page count or evaluation window supplied.
Completed example
Hypothesis: clearer titles may improve relevant search click-through. Groups: select comparable pages; retain an unchanged comparison group where feasible. Before editing: record page/query impressions, clicks, position, current titles and dates. Change: title wording only; log concurrent changes. Evaluation: agree the window and minimum data before starting, then compare like-for-like queries and groups. Limits: search changes and audience shifts can confound results; no traffic improvement is assumed.
Load this input into the prompt, then copy it to WebAct to try the task. Your result may differ from the illustration.
Decisions and troubleshooting.
Can one traffic increase prove the SEO change worked?
No. Seasonality, other edits and search changes can influence results. Interpret observations against the comparison and known confounders.
Why is the experiment impossible to interpret afterward?
Check whether multiple changes were mixed or the baseline was missing. Document those limits rather than assigning the outcome to one edit.
Try it with your own source.
Replace the example with your material in the task prompt. Keep the requirements you need, then copy the task into WebAct.
Customize and copy the task ↑