SEO Experiments based on GSC and GA4 data
Run before and after or control group SEO experiments on any page, group, or segment, all based on your Google Search Console and GA4 data.
Core features of the SEO Tests tool
Before and after tests
A before and after test compares your metrics for a set period before a change with the same length of time after it.
Pick a single page, a group of pages, or a saved segment as the scope. Sitechecker pulls the matching prior period automatically, aligned by day of the week, so the two periods are comparable.
You see the change for every selected metric across the test period and the comparison period.
Pick a single page, a group of pages, or a saved segment as the scope. Sitechecker pulls the matching prior period automatically, aligned by day of the week, so the two periods are comparable.
You see the change for every selected metric across the test period and the comparison period.
Control group tests
A control group test compares your changed pages against a set of similar pages you leave unchanged. This isolates the effect of your change from wider trends in search.
For every metric you see the change in the test group, the change in the control group, and the net lift between them. The final verdict is based on that net lift, not the raw change in the test group alone.
For every metric you see the change in the test group, the change in the control group, and the net lift between them. The final verdict is based on that net lift, not the raw change in the test group alone.
Metrics (GSC + GA4)
Every SEO experiment is built on Google Search Console data, with GA4 metrics available on top.
From GSC you can track clicks, impressions, CTR, average position, and ranking keywords. From GA4 you can add sessions, bounce rate, key events, average session duration, and session key event rate.
Choose one or more metrics, and set the country and device the test should measure.
From GSC you can track clicks, impressions, CTR, average position, and ranking keywords. From GA4 you can add sessions, bounce rate, key events, average session duration, and session key event rate.
Choose one or more metrics, and set the country and device the test should measure.
Pages and keywords
Each experiment has its own detail page with summary cards, a chart, a Pages view, and a Keywords view.
The Pages view shows how individual URLs moved between the two periods. The Keywords view shows the query-level change from Google Search Console.
GA4 metrics appear at the summary and Pages level, since they are page based, while the Keywords view stays GSC based.
The Pages view shows how individual URLs moved between the two periods. The Keywords view shows the query-level change from Google Search Console.
GA4 metrics appear at the summary and Pages level, since they are page based, while the Keywords view stays GSC based.
Experiment log
The experiment log keeps every SEO test in your project in one place. You can see who created each test, its type, its status, and its result without asking around.
Search by name and filter by status, type, scope, or the person who created it. Anyone with access to the project can open a test and review its results.
Search by name and filter by status, type, scope, or the person who created it. Anyone with access to the project can open a test and review its results.
Inside the SEO Tests tool
Scope by segment or URL list
You can set an experiment's scope with a saved segment or a list of URLs. Add URLs manually, pull them in from your project segments, or upload them by CSV.
If you use a segment, the test stays tied to that segment, so its scope updates when the segment does.
If you use a segment, the test stays tied to that segment, so its scope updates when the segment does.
Country and device filters
Each experiment can be filtered by country and device. Device supports desktop, tablet, and mobile, and you can select one or several at once.
The filters apply to your Google Search Console data the same way they do across the product.
The filters apply to your Google Search Console data the same way they do across the product.
Automatic comparison period
The comparison period is calculated for you, never picked by hand. It takes the previous period of the same length and aligns it by day of the week so the comparison is fair.
You can set an optional limit date to control how far back the comparison period is allowed to reach.
You can set an optional limit date to control how far back the comparison period is allowed to reach.
Success metric and goal
Every experiment has one success metric and a target growth set as a percentage.
When the actual change meets or beats the goal, the result is marked Success, otherwise it is Failed. If there is not enough data to judge the change, the result is marked No result instead.
When the actual change meets or beats the goal, the result is marked Success, otherwise it is Failed. If there is not enough data to judge the change, the result is marked No result instead.
Test status and progress
Each test moves through Queued, Running, Finished, and Stopped, with a progress bar based on time elapsed.
While a test runs you can extend it to add more days, or stop it early and keep the data collected so far. The detail page also shows when full data for the final result is expected.
While a test runs you can extend it to add more days, or stop it early and keep the data collected so far. The detail page also shows when full data for the final result is expected.
Completion email
When an experiment finishes, the person who created it gets an email that the results are ready. The email states the final result as Success, Failed, or No result.
No other alerts are sent, so the log stays quiet until a test is actually done.
No other alerts are sent, so the log stays quiet until a test is actually done.
FAQ
What is the difference between a before and after test and a control group test?
A before and after test compares one set of pages against their own past performance. A control group test compares your changed pages against similar pages you leave unchanged.
The control group version helps you separate the effect of your change from wider trends in search.
The control group version helps you separate the effect of your change from wider trends in search.
Do I need Google Search Console connected to run an SEO experiment?
Yes, a connected Google Search Console account is required for every SEO experiment. The whole experiment is built on GSC URLs and GSC data, even if you only display GA4 metrics.
You can connect it through the GSC Dashboard and reuse the same data across the product.
You can connect it through the GSC Dashboard and reuse the same data across the product.
Which metrics can I track in an SEO experiment?
From Google Search Console you can track clicks, impressions, CTR, average position, and ranking keywords. From GA4 you can add sessions, bounce rate, key events, average session duration, and session key event rate.
GA4 metrics show at the summary and Pages level, while the Keywords view stays Google Search Console based.
GA4 metrics show at the summary and Pages level, while the Keywords view stays Google Search Console based.
How does the tool decide if an experiment passed or failed?
You pick one success metric and set a target growth as a percentage.
For a before and after test, the result compares the test period against the automatic comparison period. For a control group test, the result is based on the net lift between your test group and control group.
If the change cannot be measured reliably, the experiment is marked No result.
For a before and after test, the result compares the test period against the automatic comparison period. For a control group test, the result is based on the net lift between your test group and control group.
If the change cannot be measured reliably, the experiment is marked No result.
Can my whole team see and manage these experiments?
Anyone with access to the project can view any experiment, open its results, and duplicate it. Only the person who created a test can edit it before it starts, extend it, stop it, or delete it.
This keeps every SEO experiment visible to the team while leaving control with its owner.
This keeps every SEO experiment visible to the team while leaving control with its owner.
How do I add pages to an experiment?
You can add pages as a saved segment or as a list of URLs. URLs can be added manually, pulled from your project segments, or uploaded by CSV.
A control group test uses a separate list of URLs that cannot overlap with the test group.
A control group test uses a separate list of URLs that cannot overlap with the test group.
