Skip to content

Lesson 2 of 3

Handling ranking drops

Diagnose a drop in an hour, tell the client before they ask, and turn it into a plan.

5 min readIntermediateUpdated 2026-08-22

The email arrives before your first coffee: a screenshot of a chart heading down and the words "what's happening to our rankings?" The two worst replies are instant reassurance and silence. Every drop has one of five causes, a tracking artefact, a change on the site, a change in the results page, a Google update, or a competitor doing something better, and you can usually tell which within an hour by checking them in that order.

First, prove the drop is real

A rank tracker takes one sample a day, from one market, and Google's results vary by location, device and searcher. A move of a few positions on a single day is noise, so read the position history over the past two or three weeks before treating a dip as an event. A day where the tracker finds the site nowhere at all deserves the same scepticism: check again tomorrow before declaring an emergency.

Then cross-check in Google Search Console's Performance report. A real drop loses impressions and clicks for the affected queries; a tracking blip doesn't. An independent second reading helps too, and the free Google rank checker gives you one, fresh and outside whatever history the tracker holds. If Search Console looks steady while the tracker fell, you have a measurement story rather than a ranking story, and you can tell the client so within the first fifteen minutes.

Second, check your own site

Self-inflicted drops usually trace to a recent deploy. Ask what changed on the site around the date the drop started, then verify the answer instead of trusting it. Fetch the affected page and confirm it returns a 200 status, still contains the content that was ranking, carries no stray noindex tag, and has a canonical pointing to itself. If the site was redesigned or migrated, trace the old URLs too: a redirect chain that ends in an error, or a page that moved with no redirect at all, explains a drop on its own.

Search Console's URL inspection tool shows the page as Google last saw it, which settles arguments about what a deploy actually did.

Third, read the results page itself

A position can hold while clicks fall, because the page around the position changed. An AI-generated overview now answers the query at the top, a map pack appeared, more ads run above the results, or the layout now favours videos. Open the query in a private browsing window and look at what sits above the client's result. If the layout changed, "recovering the ranking" is the wrong plan because the ranking never moved; the real plan is competing for whatever now takes the clicks, which might mean local work, or making the page worth citing.

The same look tells you about competitors. If someone new sits above the client, open their page and work out what earned it: a better match for the intent, fresher coverage, links and mentions they didn't have last quarter. That points to a content and authority response, the slowest fix on this list and the one clients least enjoy hearing about, which is exactly why it needs evidence.

Fourth, check whether Google moved

Google publishes its confirmed ranking updates and search incidents, with dates, on the Search Status Dashboard. If the drop lines up with a confirmed update, the diagnosis changes character: nothing on the site broke, Google re-weighed quality across the web, and sites in the same niche moved together. Check whether competitors dropped too, because shared pain is diagnostic.

Be careful what you promise next. Google's own guidance on core updates says a drop doesn't necessarily mean anything is wrong with a page, and Google publishes no recovery schedule. The honest plan is to keep improving the site's content quality and usefulness and to measure over the following months. Anyone who quotes a recovery date is guessing, so don't quote one to your client.

Tell the client before they ask

The retention move is the message the client receives before they open their own dashboard. Four parts, in order: what happened, in measured numbers; which of the five causes it is, and how confident you are; what you're doing about it; and when they'll hear from you next. If the hour ends without a firm cause, say what you ruled out. "It isn't the site and it isn't a confirmed update, so we're studying the results page" is a diagnosis in progress, not an admission of failure.

Then fold the drop into the plan rather than filing it as an incident. A site-change drop becomes a fix with a verification date. A layout change becomes a proposal. An update becomes content improvement in next quarter's calendar. Reading rank movements honestly covers the discipline of not overclaiming in either direction.

What to take away

  • A drop has five possible causes, the tracker, the site, the results page, a Google update, or a competitor, and you check them in that order because the early ones are the fastest to rule out.
  • Confirm a drop in Search Console before diagnosing it: real drops lose impressions and clicks, tracking noise doesn't.
  • A position that holds while clicks fall means the results page changed around it, and the answer is competing for the new layout, not recovering a ranking that never moved.
  • When a confirmed Google update is the cause, say so, improve quality site-wide, and never promise a recovery date, because Google doesn't publish one.

Next

A drop handled well often surfaces work the retainer doesn't cover. Upsells that help clients shows how to offer it without turning the review into a sales call.

Free tools this lesson uses

Saved in this browser only.

Chapter 6: Keep clients for years, not quarters

All 8 chapters