In the performance report, clicks and impressions can be broken down by search query, page and country. This very separation prevents a single drop from being hastily explained as a general SEO problem.
Is the drop everywhere or only on a few pages?
Compare a suitable period with the previous one or with the previous year. Sort pages by click difference. A problem on one important URL needs a different investigation than a drop across an entire directory.
Start with a clear comparison: 28 days against the previous 28 days can show a sudden break; comparing with the previous year helps with seasonal topics. Make sure the days are complete and check whether events such as a relaunch, redirects or a campaign launch fall within the period. A single week can be distorted by public holidays or waves of demand.
Then open the “Pages” table and sort by the difference in absolute clicks. A page with 5,000 instead of 7,000 clicks explains more of the overall drop than ten pages that each lost two clicks. Check impressions and click-through rate at the same time. If impressions fall, it may be about visibility or demand; if they stay stable and the click-through rate falls, the presentation or the intent behind the queries may be changing.
Mark the few URLs that explain most of the difference. Only when many groups of pages are affected at the same time does a site-wide check make sense.
Read search queries and countries separately
Open the affected pages and look at their most important search queries. Then check countries, devices and search types. This shows whether a change affects only one market, one group of demand or one type of display.
First filter one affected URL and then look at its search queries. This prevents strong topics on another page from obscuring the picture. If only queries for one particular service are losing, check that content and the search results; if your own brand is losing too, a different investigation is needed.
Then separate countries and devices. A drop in Germany with stable traffic from Austria can point to regional demand or regional changes in the results. A purely mobile problem after a design change deserves a technical check. Also compare like with like: web search, images and video should not be mixed without reason.
Write down the combination of page, query, country and device. “The domain is losing” is too broad as a task; “three service pages are losing mobile impressions for five non-brand queries” can be tested.
What a falling curve does not yet prove
Fewer clicks can be related to fewer impressions, changing search interest, seasonality, a technical problem or a different presentation of results. Google explicitly recommends checking external factors and recurring seasonal patterns as well. Average position alone is no verdict on a page.
Search Console shows search performance, but not the cause on its own. Fewer impressions can result from lower search interest, seasonal fluctuations or changed visibility. Fewer clicks with similar impressions can point to the click-through rate and the visible result environment. Average position summarises many queries and situations; interpreted in isolation, it can be misleading.
With an abrupt drop, also check technical signals: are the affected pages still indexed? Were there redirects, server errors, a wrong canonical or an accidental `noindex`? With a slow decline, it is more worthwhile to look at topics, competing answers and how up to date the page is.
Do not confuse correlation with cause. The fact that the drop began after an update does not prove that the update hit the page. Nor does a new headline published at the same time prove that it is to blame.
Choose a next step you can verify
Formulate a hypothesis with a test: “These five pages have been losing for these queries since the relaunch” can be checked. “Google doesn’t like us any more” cannot. A technical check only follows a technical pattern; if demand has shifted, a new page may be the wrong reaction.
Formulate a hypothesis with an expected signal. Example: “Since the change, five former guide URLs redirect to a generic category page; their former search queries are losing impressions.” Then check the redirects, the new target pages and the affected queries. If the pattern is confirmed, correct the mapping in a targeted way and observe the same segments again.
A second example: clicks for a seasonal service fall compared with the previous month but are above the same month of the previous year. A hectic rebuild of the page would hardly be justified here. Document the finding and check the next suitable period.
The investigation ends with three lines: affected segments, the most likely explanation with a cross-check, and a concrete measure or deliberate observation. This turns a worrying curve into a decision.
