Skip to content

Guide

Measure the effect of an SEO change in Search Console

Updated November 21, 2026

In short

Compare equal windows before and after the day a page went live, and set the page's change against the whole site's change. Wait for final data, leave out pages with too few clicks to judge, and say what moved beside the change, not that the change caused it.

Plans
Every plan, downloads included.
Credits
None. The report reads numbers PerfBee already keeps.
Needs
Search Console connected, for the clicks.

Start from the day the change went live

Measure from the day the new version was on the page, not the day the work was done. A release can wait in a queue, and a cache can serve the old page for days.

Change one thing on a page at a time. A second change inside the window mixes two effects, and neither can be read on its own.

Compare 28 days after with the 28 days before

Take two windows of the same length, so that each holds the same weekdays. 28 days is four full weeks, long enough for most pages to collect clicks worth reading.

Search Console's final data trails by about 3 days. So the 28 days after a change can be read 31 days after it went live, and not before.

Set the page against the whole site

A page whose clicks rose by a tenth while the whole site rose by a fifth did worse than the site. Compare the change of the page with the change of the site over the same two windows.

Then look at the same days a year ago, so that a season does not pass for a result.

Leave out pages with too few clicks

Small numbers swing by chance. A page that went from 6 to 9 clicks rose 50%, and that says nothing. When both windows hold fewer than 20 clicks, give the numbers and no percentage.

Read the position and the click rate as well as the clicks. A new title usually shows first in the click rate, at about the same position. A change to the content usually shows in the position.

Say what moved beside the change, not why

Search results move for many reasons at once: Google's updates, other sites, the season, the demand. A before and after comparison shows what moved next to your change. It cannot prove that the change caused it.

Write it that way in a report. Name the clicks before and after the page went live, and how far the whole site moved in the same days.

Read it in Impact Reports

Impact Reports covers the last 28 days, 90 days or 180 days, against the days before and the same days a year ago. It lists the fixes a crawl confirmed, the ones that came back and each rewritten page.

For each page that AI Page Rewriter rewrote, the clicks are for the 28 days before and the 28 days after the page went live. The page's change stands beside the site's change. AI Page Rewriter is on Business and up.

A page shows Waiting until its days are complete, and Too few clicks under 20 clicks in both windows. It shows Measured once it can be judged, or Not measurable when the numbers cannot be compared.

Download this report gives a PDF or Excel on every plan, with your logo and name on Starter and up: the fixes confirmed, the pages rewritten and the clicks before and after, ready to send.

Measuring a change, answered.

Anything else? Talk to us.

How long should I wait to measure an SEO change?

At least 28 days after the page went live, plus about 3 days for the final Search Console data. A shorter window holds too few clicks for most pages.

Can Search Console prove that my change worked?

No. It shows clicks, impressions and positions before and after. Set the page against the whole site and against the same days a year ago, and report what moved beside the change.

Why compare a page with the whole site?

The whole site moves with the season, the demand and Google's updates. A page that rose less than the whole site did worse than the site, even when its clicks went up.

What if a page has very few clicks?

Give the numbers and no percentage. PerfBee shows Too few clicks when both windows hold fewer than 20 clicks.

Should I also compare with the same days a year ago?

Yes. The days right before can be a different season. The same days a year ago meet the same season, so a seasonal rise is not read as the change working.

Show what your work did.

Open Impact Reports in the site menu for the fixes a crawl confirmed and each rewritten page set against the site.

Start for free

Then try Business free for 14 days, with no card.