Diagnostic playbook for a drop with no apparent cause

When you have already ruled out a manual action, already checked Google's update calendar and found nothing, and the site has not gone through any recent technical change, the next step is to deepen the granularity of the analysis until something shows up, instead of giving up or starting over from scratch.

This is the most common scenario within the universe of traffic drops reported in SEO forums: sites that ruled out the obvious causes and remain without an answer, asking other professionals what they would investigate first. This article is that playbook.

It assumes you have already read the diagnostic guide to Google traffic drops and already gone through the initial triage. If you have not checked for a manual action yet, stop here and go back to the diagnostic guide first.

In practice, "no apparent cause" almost always means one of three things:

1. The cause exists, but it is hidden at a level of granularity that the general report does not show (a page category, a country, a device type).
2. The cause is a combination of small factors that, added together, produce a large effect (a drop in authority for a set of pages, plus a competitor that rose, plus a subtle shift in search intent).
3. The cause is technical, but subtle enough to pass a surface level check (an error in part of the sitemap, a robots.txt block that affects only one directory, a template change that altered the main content by accident).

In practice: it is almost never "the algorithm decided to hurt me". It is almost always something specific that has not yet been isolated.

The Search Console performance report, viewed in aggregate, hides patterns. Segment by page, by query, by country and device, and by search type, as shown in the summary below.

If the drop is concentrated in a specific segment, you already have a concrete lead instead of a vague suspicion about "the algorithm".

Build a simple list, with dates, of everything that happened on the site and around it in the last 90 days: deploys, template changes, a CMS or plugin switch, an agency change, paid campaigns that started or stopped, price or catalog changes (in e-commerce), and any market event in your niche.

A case reported in SEO forums illustrates the problem well: a professional listed four possible changes (the client's address change, a Google update in the same window, a spike in spam backlinks, and a PageSpeed Insights score that fell from 72 to 38) and concluded, correctly, that there was no single obvious culprit: the drop was probably the result of the combination of all four factors, not one isolated one.

Crossing this timeline with the drop curve, day by day, usually reveals coincidences that memory alone would not recall.

If everything inside the site looks normal, the next place to look is outside it:

• Competition: did someone in your niche gain position exactly on the queries you lost? Use a position tracking tool to confirm it, not just the impression.
• SERP change: did the results page for your main queries gain a new feature (a panel, a carousel, an AI generated answer) that pushed organic results down? This specific pattern, of stable impressions and falling clicks, has its own playbook in AI Overviews and the CTR drop.
• Seasonality and demand: is there a market reason (season, news, a change in behavior) for search for your topic to have genuinely dropped, independent of your site?

A page can show up as "indexed" in the coverage report and still not actually be competing for the queries that used to bring it traffic. It is worth inspecting a sample of affected URLs directly in the Search Console inspection tool, and checking whether Googlebot is crawling the site at the expected pace. If the pattern points to a crawling problem, the topic is covered in more depth in crawl budget and Googlebot behavior.

After going through the four steps, two things can happen. Either you find a dominant factor and can act on it, or you end up with a list of two or three plausible factors, none of them obvious on its own.

This second scenario is frustrating, but it is not uncommon, and it does not mean the diagnosis failed. In these cases, the strategy changes: instead of looking for "the cause", the work becomes fixing the factors you can control (performance, the quality of a content segment, backlink profile) and monitoring whether traffic responds in the following weeks.

If I had to start with just one point, I would pick step 1: segment the performance report by page, then by query, then by country, noting where the drop concentrates. That alone already solves a good share of cases that seem to have no cause. The full picture of causes is in My traffic dropped: a diagnostic guide to Google traffic drops, and the cases of manual action and core update have their own playbooks.