Was that traffic drop real, or a Search Console data anomaly?

A Search Console traffic drop is more likely real when clicks fell, because Google says its documented impressions logging error did not affect clicks. If only impressions, click-through rate or average position moved, a reporting error is plausible. Either way, check Google's data anomalies page for your dates, then measure impressions against clicks.

• Google's data anomalies page records known issues from the last 3–16 months, and its newest entries were dated 13 August 2026 when we read it (Google, Search Console Help, read 30 October 2026).
• Google says an impressions logging error ran from 13 May 2025 until 27 April 2026, did not affect clicks, and may show as a decrease in impressions (Google, Search Console Help, read 30 October 2026).
• Google says collected Search Console data is usually available in 2–3 days (Google, Search Console Help, read 30 October 2026).
• Penlid's anomaly check compares the change in impressions with the change in clicks, and its chance band is within ±10 points only above about 770 clicks in each 28-day window (Penlid, Search Console anomaly check, read 30 October 2026).

Does Google publish a list of Search Console data errors?

Yes: Google's help page, Data anomalies in Search Console, lists known issues in the last 3-16 months that might affect your data. Google's debugging guide also names a change in the data processing or a logging error among the causes of a drop in organic Search traffic. The table summarizes the page's entries as Penlid's Search Console anomaly check copied them on 30 September 2026; a re-read on 30 October 2026 found no newer entry.

System annotations on the chart are Automatically generated by Search Console to show issues with data processing and reporting, but they won't show up in comparison mode or 24-hour views, so a Compare view with no marker proves nothing; open the plain chart for your dates.

Which ordinary effects can look like a drop?

Three ordinary effects can look like a drop with no anomaly behind them: preliminary days, differences in timing between tools, and a query filter.

• Preliminary days. Google says the newest data can be preliminary and draws it as a dotted line, so a fall confined to that part may not be final; see why the last few days in Search Console change.
• Timing. Google says collected data is usually available in 2–3 days, and the Performance report labels daily data in Pacific Time except in the 24-hour view, so a one-day dip in another tool's report can land on a different day here (our reasoning).
• A query filter. Anonymized queries are included in the chart totals unless a query filter is applied, so adding one can lower the totals with no change in traffic. Remove it before you compare two periods.

How do you tell a logging step from a real fall in clicks?

Compare the change in impressions with the change in clicks across the same dates: a step in impressions alone fits the logging error, and a fall in clicks as well does not. Google says clicks were not affected by the error and that you may notice a decrease in impressions. That is a pattern, not proof, because a real change can move impressions and clicks apart in the same way. A fall in clicks can also be a reporting error: the Discover and job-listing entries in the first table cut clicks too, so match your dates against them first.

The anomaly check turns the pattern into a rule: the change in impressions minus the change in clicks, in percentage points, across 27 April 2026, the day Google's fix lands. It uses the 28 whole days before and after, with that day set aside; 28 days is four whole weeks, so every weekday appears four times per side. The 10-point line is the tool's own choice, not a cliff in the evidence.

To measure across 27 April 2026, the file needs one row per day for 30 March to 26 April and for 28 April to 25 May 2026, with clicks and impressions both on, running to 26 May 2026 or later because the newest day is set aside. It also needs enough clicks: the chance band is within ±10 points only above about 770 clicks in each 28-day window, about 27 a day. Our arithmetic reproduces that figure: the change between two equal Poisson counts of n clicks has a 95% band of about ±1.96 × √(2 ÷ n), and 1.96 × √(2 ÷ 770) = 0.0999, about ±10 points. For a drop on other dates the measurement does not apply, but the list of overlapping anomalies still does.

What should you check once no anomaly explains the drop?

Check whether the drop is a season, an update or a change confined to part of your site. Google's debugging guide suggests comparing the drop period with a similar one, for example Compare last 3 months year over year, and checking Google Trends to understand if the drop was only for your website or throughout the web. Our guide to seasonality in SEO data covers the first check.

Then look for ranking updates on the Google Search Status Dashboard, whose ranking table included the May 2026 core update (from 21 May) and the September 2026 spam update (from 24 September). For a core update Google recommends waiting at least a full week after a core update completes before analyzing your site in Search Console; our guide to measuring what a Google update did to your site treats the result as timing evidence, not proof. To see whether a real fall sits on one page, a group of pages or the whole site, read content decay versus a site-wide dip.