How to measure SEO impact: did your change work?

To measure SEO impact, write down what the change should do before you make it, start the clock on the day Google rereads the page, wait as long as that kind of change needs, and compare the result to what the same search would have done without you. Skip any one of those and a normal month of drift will look like your change worked, or failed.

Most teams check one number a week after deploy and call it. That reading answers almost nothing: Google may not have seen the change yet, and rankings move by a few places a month with nobody touching the page. The method below is how you tell whether your SEO change worked, using nothing but your own Search Console data.

TL;DR:

• Write the prediction first: which search, which number, by how much. A result you did not predict can't fail, so it can't teach you anything.
• The clock starts at Google's recrawl, not at deploy. Until Google rereads the page, no searcher has seen your change.
• Wait the window the change needs: about four weeks for a title, longer for a redirect or a link, up to three months for rewritten copy.
• Judge each change on the search it was made for, and only call a ranking move when it is bigger than normal drift for that position.
• Search Console stops at the click. It can show a change won more clicks, never that it won more signups.

What does "SEO impact" actually mean for one change?

It means one search, one page, and one number that moved because of what you did. Not site traffic, which rises and falls with seasons and Google updates, and not a page average, which shifts whenever the mix of searches on the page shifts. A change is aimed at a search, so that is where its impact is read.

Pick the number by what the change touched. A title and meta rewrite on a page that already ranks should move click-through rate on the searches it ranks for. Rewritten copy aimed at a search should move position on that search. A fix for an indexing problem should get the page indexed and showing at all. Writing the number down first is what keeps you honest later, because "something improved" is always true somewhere.

Did my SEO change work, or is it just drift?

You can only tell by comparing it to what the search would have done anyway. Rankings move with nobody touching the page, and the move is larger the further down the page starts. On our own data, a search ranking 10 to 20 moved a median of 2.7 places in a month untouched, and one past 20 moved 5.6.

That is the trap. In our measurements of how much rankings fluctuate, 37% of untouched searches starting between 12 and 16 climbed three places in a month on their own. So a climb from 14 to 11 after a rewrite is not evidence the rewrite worked. It may have. It is the same move a third of untouched searches made anyway.

Two habits fix most of this. Ask for a move bigger than drift: on our data, about one untouched search in ten moved half its starting position in a month, so 14 has to reach 7 before we call it. And check what the rest of the site did over the same weeks. If every page climbed, the climb belongs to Google, not to your edit.

When do you start measuring an SEO change?

On the day Google recrawls the page, not the day you deploy. Until Google rereads it, nobody searching has seen the change, so every day before the recrawl is measuring the old page. On a small site that gap can be days or weeks.

Search Console's URL Inspection tool shows the last crawl date for any page. Note the deploy date, then check the page until the crawl date passes it. That is day zero. Requesting indexing in the same tool can shorten the wait, but it does not guarantee a crawl.

How long should you wait before judging?

As long as the change needs to settle, which depends on what it touched:

• A title or meta description: about four weeks. Google shows the new title soon after the recrawl, and you need a few full weeks of impressions to read click-through rate.
• A redirect, canonical, sitemap or internal link: about six weeks. Google has to recrawl more than one page before the change counts.
• Rewritten copy aimed at a search: up to three months. Google's own guidance is that quality changes can take months to be fully reprocessed.

Read whole weeks, never a few days. Weekends alone can look like a 30% drop on a business site. And do not edit the same page again inside the window: a second change resets the baseline, and neither one can be measured.

Which Search Console number should you read?

Clicks and impressions at page level, position at query level. Page-level totals are close to complete. Query-level rows under-count, because Google hides rare searches, but they are the only place a position for one search lives. That split is why page-level and query-level data answer different questions.

Position is an average weighted by impressions, so read it with its volume. A position of 4 on 12 impressions is a guess. Under about 30 impressions a day, ignore single-day positions and read the week.

How do you measure SEO impact without doing it by hand?

Keep a log: the search, the page, the date of the change, the recrawl date, the prediction, and the date it can be judged. Then read the result against drift when that date comes. It works, and it is tedious enough that most teams stop after the second change.

That log is what QueryScope keeps for you. It tracks each SEO change from the day Google rereads the page: your agent records the change and its prediction from the IDE, the clock starts at the recrawl, the search is not handed back for another edit while it settles, and the result is read against what the rest of the site did over the same weeks. When the window closes, the prediction either held or it did not.

What can't Search Console tell you about SEO impact?

Anything after the click. Search Console sees impressions, position and clicks. It never sees whether the visitor signed up, bought, or left. A change that doubles clicks to a page that does not convert has an SEO impact and no business impact, and Search Console cannot tell those apart.

So report SEO impact as what it is: more of the searches you chose to win, measured honestly. What those visitors do next belongs to your analytics, and to a different question.

Sources

• Google Search Console Help, URL Inspection tool (last crawl date and requesting indexing).
• Google Search Central, Google Search's core updates and your website (improvements to content can take months to be reflected).
• Google Search Central, A deep dive into Search Console performance data filtering and limits (rare queries are left out of the query table while their clicks stay in the totals).
• QueryScope's own measurements of untouched ranking drift on five connected sites, described in why Google rankings fluctuate.