This article was AI-generated as part of an experimental historical-content project. The date reflects the period being analyzed rather than the date the article was originally written.
The conventional rhythm around a Google core update is that the week it lands is the week to worry. Rankings move, tracking tools light up, forums fill with screenshots of traffic charts, and somebody in the company asks what just happened to the results for the brand name.
I think that rhythm has it backwards. The most informative moment of a core update is not the start. It is the end, and then a week after that.
The rollout is finished
Google’s Search Status Dashboard now says of the March 2025 core update: “The rollout was complete as of March 27, 2025.” It began on March 13, so it took about the two weeks Google estimated. Search Engine Land reported big swings in both directions for some sites, while noting that the update did not seem as widespread as some recent ones.
Earlier this month I wrote about the attribution problem this update creates, since Google also promised a separate series of creator-focused improvements through the year. Completion does not solve that. It does solve a simpler problem that gets less attention, which is timing.
Why mid-rollout readings mislead
A core update does not switch on all at once. Results can move, settle and move again while the change propagates. Barry Schwartz’s recap at Search Engine Roundtable describes that pattern this time: movement starting around March 16, a calmer stretch, confusion in some third-party tracking tools a few days later, and another burst of volatility around March 26, the day before Google called it done.
A snapshot taken on March 18 captured one of those intermediate states. If a news article about your CEO jumped onto the first page that week, or your own careers page dropped off it, there was no way yet to know whether that was the new arrangement or a stop along the way. Decisions taken from that snapshot, such as rewriting pages, commissioning content or escalating to leadership, were taken on partial data.
Google’s own guidance says the same. Its documentation on core updates tells site owners to confirm that the rollout has finished, and recommends “waiting at least a full week after a core update completes before analyzing your site in Search Console,” then comparing that week with a week before the rollout began. For this update, that means the first fair comparison is roughly the week starting April 3 against the week of March 6.
Brand results need a different instrument
Search Console has a limit that matters for reputation work: it only measures pages you own. The results for a company or an executive name are mostly pages you do not own. News coverage, Wikipedia, review sites, LinkedIn, forums, sometimes a court filing. The update may have changed how Google weighs every one of them, and none of that appears in your reports.
So the diagnosis has two halves. The first is Google’s method, applied to your own pages, with like-for-like weeks. The second has to be done by snapshot: capture the first page for the names that matter, signed out and in your main markets, and compare it with a capture from before March 13 if you have one. Record what kind of source holds each position, not only which URL. A news article replaced by a forum thread in position four is a different story from one news article replaced by another. And check the AI Overviews for those same queries, because a change in which pages rank can change which pages the overview cites.
Patience as method
The temptation is to act while the movement is visible and the anxiety is fresh. But acting during a rollout means acting on noise mixed with signal. The useful question for the next two weeks is narrower: now that Google says the change is complete, what actually stayed different?
That can only be answered once the movement stops. As of yesterday, it has. By next Thursday, it can be measured.