March 2026 core update

Three updates in five weeks

Google shipped three separate algorithm events in quick succession, each affecting a different part of search. Identifying which one moved your traffic is step one, and skipping it sends the remediation in the wrong direction.

Update	Dates	What it affected	Measurement (how to tell it was this one)
February 2026 Discover core update	February 5 to 27, 2026	Google Discover feed only, not web search	Search Console Discover report fell, web search held
March 2026 spam update	March 24 to 25, 2026	Spam policy enforcement	Loss concentrated on pages matching a spam policy pattern
March 2026 core update	March 27 to April 8, 2026	Broad content quality reassessment across search	Site-wide or template-wide loss beginning March 27 or later

If your Discover traffic fell in February and you treat it as a core update problem in April, every fix you make will be aimed at the wrong system. Switch Search type to Discover in Search Console to separate the two before doing anything else.

What Google said

The dashboard note was terse: Google released the March 2026 core update and the rollout might take up to two weeks. No accompanying blog post, no stated goals, no named focus areas. Google classified it as a regular update and said nothing further.

What the data showed

• Query intent satisfaction was weighted more heavily. Why: Pages that matched a search term but did not satisfy the underlying reason for the search were demoted more reliably than before.
• Information gain became a clearer differentiator. Why: Google appeared to evaluate how much genuinely new information a page adds compared with what already ranks. Rephrasing the existing top ten stopped being sufficient.
• The update was widely described as moderate. Why: Practitioners who compared it with the May 2026 core update that followed generally found March the milder of the two, which matters when you are attributing a drop across both windows.

How to measure your impact

Google recommends at least a week after completion before drawing conclusions. With the rollout ending April 8, the earliest reliable analysis date was around April 15. Compare March 1 to 23 as your baseline against April 9 onward.

• Sort pages by average position change, not by traffic loss. Why: Traffic follows position. Sorting by position decline surfaces the pages Google actually re-scored.
• Export the top 50 losers and look for the common thread. Why: Same content type, same topic cluster, same author, or no author at all. That pattern is your diagnosis, not a list of individual page problems.
• Check Discover separately. Why: A February Discover loss and an April core loss look identical in a combined view and need completely different fixes.
• Do not make major changes before you know which pages need them. Why: Editing pages that were unaffected or that improved is how a recoverable drop becomes a larger one.

For the diagnosis sequence in full see the Search Console recovery guide, and for what to do in the first two days after any update lands, the first 48 hours checklist. The core update recovery service covers the recovery method.