Search Console Data Quirks to Know When Analysing Updates

Google Search Console is the best free source of truth you have when an algorithm update lands. It shows real clicks and impressions from Google, not estimates. But it also has quirks that can make a normal week look like a disaster, or hide a real problem behind a comforting average.

If you analyse an update without knowing those quirks, you can easily blame the wrong cause. A falling average position might mean you are ranking for more queries. A sudden impression drop might be a logging issue Google has already documented. Missing queries might just be anonymised.

This guide walks through the Search Console data quirks that matter most during update analysis in 2026, what Google's documentation says about each one, and a step-by-step method for getting a reliable before-and-after picture.

Key Takeaways

• The newest data in the Performance report is preliminary and can change; Google recommends waiting at least a full week after a core update completes before analysing.
• Average position is an average of your topmost position across queries with impressions, so it can fall even when traffic grows.
• Chart totals and table rows can be aggregated differently (by property or by page), so numbers do not always add up.
• A large share of queries can be anonymised; one 2022 Ahrefs study reported 46.08% of clicks went to hidden queries.
• Always check Google's Data anomalies page before blaming an update for a sudden change.

Why Search Console data needs careful handling during updates

Most core updates listed on Google's Search Status Dashboard since late 2021 took roughly two to three weeks to roll out, though the range runs from about six days (December 2024) to 45 days (March 2024). During that window rankings can move up and down several times. Search Console records all of that movement faithfully, but it also aggregates, filters and occasionally mis-logs data. Your job is to separate real ranking change from reporting artefacts.

If you are new to the report, our Search Console Performance report guide covers the basics. This article focuses on the parts that trip people up during update analysis.

The quirks at a glance

Quirk	What happens	How it can mislead you	What to do
Preliminary data	The newest data is still being collected and can change	The last day looks like a sharp drop	Exclude the most recent days from comparisons
Average position	Averaged across all queries with impressions	New low-ranking queries pull the average down	Analyse position per query and per page
Aggregation by property vs page	Chart and query data aggregate by property; page data by page	Totals don't match between views	Compare like with like
Anonymised queries	Some queries are hidden for privacy	Query totals fall short of page totals	Use page-level data for traffic, query data for direction
Top rows only	Search Console stores top data rows, not all rows	Long-tail losses disappear from tables	Use the API or bulk export with filters
Logging anomalies	Google occasionally documents reporting errors	A data issue looks like an update hit	Check the Data anomalies page

Quirk 1: Data delays and preliminary data

Google's Performance report help page says: "The newest data can be preliminary, meaning it's still being collected and might change in the next few hours." Preliminary data is shown as a dotted line on the chart. During an update, people often screenshot the last day, see a dip and panic. That dip may simply be incomplete.

Google's core update guidance goes further and recommends a wait before you analyse at all:

“We recommend waiting at least a full week after a core update completes before analyzing your site in Search Console.”

— Google Search Central, Google Search's core updates

The same page suggests comparing a full week after the update with a week before it started, which keeps day-of-week patterns aligned. Our post on diagnosing a core update drop in Search Console shows that comparison step by step.

Quirk 2: Average position is not your "ranking"

According to Google's metrics documentation, position is "the topmost position occupied by a link to your property or page in search results," and it is "averaged across all queries in which your property appeared." A position is only recorded when the result gets an impression.

That has two practical consequences during updates:

• Average position can fall while traffic rises. If an update lets you appear for hundreds of new queries at positions 30 to 60, your average gets worse, even though your existing rankings may be unchanged.
• Average position can improve while traffic falls. If you lose impressions for many low-ranking queries, the remaining set looks better on average.

Always break position down by query and by page, and read it alongside impressions and clicks. Our guide to query-level analysis after an update explains how.

Quirk 3: Property versus page aggregation

Google's Performance report help says data on the chart "is aggregated by property", and table data grouped by queries, countries, devices or dates is also aggregated by property, while data grouped by pages or search appearance "is aggregated by page." When several of your URLs appear for one query, property aggregation counts one impression with the best position, while page aggregation counts each URL.

The result: totals in the Pages tab often exceed totals in the Queries tab. During an update, compare the same view before and after, and do not mix tabs in one calculation.

Quirk 4: Anonymised queries and stored row limits

Search Console hides some queries to protect user privacy. How many? In July 2022, Search Engine Roundtable reported an Ahrefs study of 146,741 websites that found 46.08% of clicks were attributed to anonymised queries, with some sites above 80%. The same report noted Google's help text that "due to internal limitations, Search Console stores top data rows and not all data rows."

For update analysis this means:

• Query tables are a sample. Use them to see which kinds of queries moved, not to total traffic.
• Long-tail losses can be invisible in the interface. The Search Console API and bulk data export can return more rows than the interface tables.
• Filtering by page and then by query gives a more complete picture for priority URLs.

Regex filters help you group queries meaningfully, for example branded versus non-branded. See our Search Console regex guide and the post on filtering branded queries during update analysis.

Quirk 5: Documented logging errors

Google keeps a Data anomalies in Search Console page. Its introduction reads: "On rare occasions, there might be an event in Search Console that could affect your report data." Entries listed there in 2026 show how much this matters:

• A logging error "prevented Search Console from accurately reporting impressions from May 13, 2025 until April 27, 2026." Only impressions, CTR and average position were affected; clicks were not.
• Two days, 28 February and 1 March 2026, are missing from Bulk data exports for some properties, and Google says that data won't be recovered.
• As of 7 May 2026, FAQ rich results stopped appearing in Google Search, so reported FAQ impressions dropped.
• Several 2026 logging errors lowered Discover clicks and impressions on specific dates.

The impressions issue is a good example. Anyone comparing impressions from mid-2025 against mid-2026 could see a "drop" that has nothing to do with any update. Clicks were the safer metric for that period. Our post on tracking errors versus real ranking drops covers more examples.

Quirk 6: AI features are blended into the totals

Google's AI features documentation says sites appearing in AI Overviews and AI Mode "are included in the overall search traffic in Search Console" and are reported in the Performance report. You cannot assume that a change in clicks came from classic blue-link rankings alone. If impressions hold steady but clicks fall, a change in how results are displayed may be part of the story, which is why rankings and traffic impact should be assessed separately.

A step-by-step method for clean update analysis

1. Confirm the dates. Note the start and completion dates from the Search Status Dashboard.
2. Check Data anomalies. Rule out a documented logging issue for the same period.
3. Wait. Allow at least a full week after completion, as Google recommends.
4. Compare equal periods. A full week (or 28 days) after versus the same length before the rollout started.
5. Segment. Split by search type (Web, Discover, News), device, country, and branded versus non-branded queries.
6. Go page-first. Find the pages with the largest click changes, then look at their queries.
7. Read position per query. Separate small slips (for example 2 to 4) from large drops (4 to 29), which Google treats differently in its guidance.
8. Annotate. Record update dates and your own site changes so future comparisons have context.

For a practical way to keep that record, see our article on annotating updates in your reporting.

Which metric to trust for which question

Question	Best metric	Watch out for
Did I lose traffic?	Clicks, by page	Seasonality and preliminary days
Did rankings change?	Position per query (not site average)	Low impression queries are noisy
Did visibility change?	Impressions	The 2025–2026 impressions logging error
Did SERP layout change?	CTR at stable positions	Mixing branded and non-branded queries
Which content type was hit?	Clicks grouped by URL folder or template	Page vs property totals

Common mistakes

• Reporting the site-wide average position to clients. It hides more than it reveals.
• Comparing different lengths of time. Seven days against 28 days distorts every metric.
• Adding page and query totals together. They are aggregated differently.
• Ignoring Discover. Discover had its own core update in February 2026, so check the Discover report separately.
• Skipping the anomalies page. It takes a minute and can save weeks of misdirected work.

For broader recovery planning, our ranking drop recovery guide picks up where this analysis ends.

Related Guides

• What Google Means When It Says There's 'No Fix' for Core Updates
• Why Update Impact Looks Different Depending on Location and Device
• Not Every Traffic Change Is an Update: SERP Layout Changes Explained

Frequently Asked Questions

How long should I wait before analysing an update in Search Console?

Google recommends waiting at least a full week after a core update completes, then comparing a week after with a week before the rollout started.

Why did my average position drop while clicks went up?

Average position is averaged across all queries with impressions. Gaining many new low-ranking queries pulls the average down even if your main rankings are stable.

Why don't my query totals match my page totals?

Query and chart data are aggregated by property while page data is aggregated by page, and some queries are anonymised for privacy.

Could a sudden drop be a Search Console bug?

Yes. Google lists known reporting issues on its Data anomalies page, including a 2025–2026 impressions logging error. Check it before drawing conclusions.

Are AI Overview clicks shown separately?

Google says traffic from AI features is included in overall search traffic in the Performance report, so treat it as part of the totals.

Conclusion

Search Console is reliable when you respect its limits: wait for final data, compare equal periods, analyse by page and query, and check the anomalies log. Do that and your update analysis will point at real causes rather than reporting artefacts. If you would like a second pair of eyes on your data, our SEO team can help, or you can request a quote.

References

1. Google Search Central: Google Search's core updates
2. Search Console Help: Performance report (Search results)
3. Search Console Help: What are impressions, position, and clicks?
4. Search Console Help: Data anomalies in Search Console
5. Search Engine Roundtable: Google Search Console anonymizes tons of queries
6. Google Search Central: AI features and your website
7. Google Search Status Dashboard: Ranking history