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.
Google began rolling out the March 2026 core update at 2:00 this morning, Pacific time. The dashboard note is brief: “Released the March 2026 core update. The rollout may take up to 2 weeks to complete.” There was no companion blog post. Search Engine Journal notes that it is the first broad core update of 2026, and that it arrives two days after the March spam update finished.
On its own, a core update is routine. This week’s sequence makes a better case study, because it quietly breaks the method most teams use to read one.
The method
Google’s core update documentation gives sensible advice. Once a rollout has finished, wait at least a full week, then compare performance with a period before the update began. The idea is a stable “before” against a stable “after.”
Why this week breaks it
Count back a week from this morning and you land on March 20. The March spam update was released on March 24 and completed the next day. So the natural “week before” this core update contains another Google update in its final days. Anyone who uses that window as a baseline is comparing against a period that was itself moving.
The cleaner choice is to go back further, to the week before March 24. That brings its own problem: two separate changes now sit between baseline and result, and dates alone cannot split them.
Some companies have a third layer. In February, Google ran a core update scoped only to Discover, from February 5 to 27, which Search Engine Journal describes as the first time Google publicly labeled a core update that way. A publisher or consumer brand that saw Discover traffic shift last month should keep that out of any reading of this one, which is a broad Search update.
What the stack means for reputation teams
For a publisher, a muddled baseline produces a muddled traffic report. For a company watching its own name, it can produce something worse: a confident explanation of why a critical article climbed or a product page fell, with the wrong cause attached. A spam update acts on sites that violate Google’s policies. A core update reassesses quality broadly. They point to different responses, and mistaking one for the other leads to the wrong one.
Generative surfaces add a variable. Google has said AI Overviews and AI Mode rely on its core ranking systems, so during the rollout the sources cited for your name may change along with the blue links. Capture both.
A workable approach
Use dates as a first filter, not as an answer. Keep two reference points: the week before March 24, and whatever you captured yesterday. Movement between those two belongs to the spam update or to ordinary churn. Movement that appears from today, during the rollout, is more likely the core update, though not certainly. Alongside both, keep your own log of press coverage, site changes and Wikipedia edits, because a results page about a company moves for reasons Google never announces.
Google estimates up to two weeks. December’s broad update took 18 days. Either way, a reliable reading arrives in late April at the earliest, and it will be less certain than usual. Saying so internally, before anyone asks for a verdict, is part of the job.