Was it a Google update?

Google publishes every confirmed ranking update, with start and end dates, on the Search Status Dashboard's ranking history. That page is the only authoritative calendar; SEO news sites report suspected updates too, but the dashboard is what Google actually confirmed. Google's own advice for a suspected core update hit is to "check the Search Status Dashboard and take note of the start and end date of the core update," then in Search Console "try comparing this week with a week before the core update started rolling out."

The confirmed windows since mid-2025, as the dashboard listed them on July 11, 2026:

A drop that starts inside one of those windows is a candidate, not a conviction. A drop that starts weeks away from every window is telling you to look elsewhere: a site change, a technical problem, seasonality, or the clicks-without-rankings pattern. Google's debugging guide for traffic drops sketches the shapes: sharp site-wide steps from updates or security issues, gradual slides from technical problems or fading interest, and the repeating curve of seasonality. Compare year over year before you blame an algorithm for what the calendar does every year.

In Search Console's Performance report, set the comparison Google suggests: the week before the rollout began against a week after it completed. Then look at where the loss lives. A core update reassesses the site broadly, so its signature is a haircut across many pages and many queries at once. A loss confined to one page or one cluster, with the rest of the site flat, points at something narrower: a stronger competitor on those queries, a content problem on those pages, or a technical issue in that section.

Check the size of the position moves too. Google's guidance distinguishes a small slip, position two to four, from a large one, and says a small decline needs no drastic action. Falling from position 2 to 4 halves your clicks and means Google still likes the page; falling from 4 to 29 means the page effectively left the results. Those are different problems, and only the second one justifies real work.

One more gate before you conclude "update": open the Manual Actions report in Search Console. A core update is not a penalty and never appears there. If that report is empty, nobody demoted you by hand, and anyone selling you "penalty removal" for a core update is selling a service for a thing that does not exist. There is nothing to remove. There is only content to make better.

Export your queries for the comparison window and sort them into rough classes: brand searches for your site by name, informational questions, and the commercial or tool-like queries where the searcher needs to do something. Each class failing has a different meaning. Brand queries holding while informational queries collapse says your audience still wants you and the algorithm rerouted the strangers. Everything falling together, positions included, is the core-update signature. Informational clicks falling while impressions and positions hold is not a ranking event at all; that is the AI Overviews pattern, and it needs a different response than a rebuild.

Google's core-update documentation is unusually plain, and it is worth taking at face value instead of filtering it through people selling recoveries. Pages that fall in a core update are not necessarily broken. Google's analogy is a refreshed list of top restaurants: the ones "that move down aren't necessarily 'bad'; there are just other restaurants that make your top 20." There is no fix to apply, because nothing was flagged. The guidance is to "focus on making changes that make sense for your users and are sustainable in the long term," to self-assess the site as a whole against people-first content questions, and it adds that "deleting content is a last resort," only for what cannot be salvaged.

On timing, Google says improvement can take months: "it could take several months for our systems to learn and confirm that the site as a whole is now producing helpful, reliable, people-first content," and "if it's been a few months and you still haven't seen any effect, that could mean waiting until the next core update." It also says outright that "there's no guarantee that changes you make to your website will result in noticeable impact in search results." That is the honest frame: recovery is measured in update cycles, the confirmed cores of the past year landed roughly two to five months apart, and the work is content work, not a technical trick.

We will not dress that up. If the diagnosis says core update, the calm response is to decide, with open eyes, whether this site earns months of content investment on those terms. That decision has its own framework at pivot or persist, and if search demand for your topics still exists, the rebuild case is real. What does not work is panic-pruning, mass rewrites aimed at an imagined penalty, or paying someone to "remove" one.