How to respond to a Google core update
A few times a year, Google announces a core update, the industry holds its breath for two weeks, and a lot of sites make expensive decisions based on three days of data. This post is the calmer version of that fortnight: what a core update is, how to establish whether it touched you, and what a response looks like when it is based on evidence rather than adrenaline.
What a core update is and is not
A core update is a broad re-evaluation of Google's ranking systems, rolled out over days to weeks and announced on the Search Status Dashboard. It is not targeted at specific sites, it is not a penalty, and there is no manual action to appeal. When rankings drop in one, nothing on the site broke. The systems simply re-weighed relevance and quality across the whole web, and other pages now score better for the queries you used to win.
That framing matters because it dictates the response. A penalty has a cause to find and fix. A core update has a verdict to understand. Google's own guidance is consistent on this point: there is nothing specific to fix, and the path forward is making the site genuinely better for the queries it wants to win.
First, establish what actually happened
Wait for the rollout to finish before concluding anything. Positions swing during the rollout window and sometimes swing back. Then compare clean windows in Search Console, a stable stretch before the update against a stable stretch after, and segment before averaging: brand versus non-brand queries, page groups, countries. Sitewide averages hide the story every time.
Two false positives are worth ruling out. A visibility drop that coincides with an update is sometimes actually a SERP layout change, where position held but new furniture absorbed the clicks. And a drop concentrated in one page group sometimes turns out to be a plain technical regression that shipped the same week. Check both before attributing anything to the update.
If you lost
The honest news first: recovery is slow by design. Google has said that improvements may take months to register and that meaningful recovery often arrives only around subsequent updates, when the systems re-evaluate at scale. Anyone promising a quick fix after a core update is selling the weather.
The productive work is comparative. Take the queries that moved, look hard at what ranks now, and ask what those pages do that yours does not: tighter intent match, more genuine experience and evidence, better coverage, less boilerplate. Then improve steadily: consolidate weak overlapping pages, add what the winners have and you lack, and resist the urge to rewrite everything at once, which destroys your ability to learn what worked.
If you gained
Gains are information too. Something about the site aligned with what the systems now reward, and the winning pages are worth studying with the same rigor as losses. Keep shipping in the same direction, and bank the new baseline in your reporting before anyone forgets where you started.
The boring playbook
Annotate the update dates in every report and dashboard, permanently.
Judge nothing until the rollout is officially complete.
Segment the damage or the gain before diagnosing it.
Rule out SERP layout changes and same-week technical causes.
Improve the pages that lost by studying the pages that won.
Expect recovery on the scale of months, and say so upfront to stakeholders.
Sites that treat core updates as quarterly weather and keep investing in substance tend to compound while their competitors thrash. The update is not the strategy. It is the scoreboard.