A headline number like “organic clicks down 25%” tells you that something happened. It does not tell you what. Two sites with the same drop can have completely different stories: one lost a handful of high-volume informational queries to AI Overviews, the other slipped from position 3 to position 8 on dozens of money keywords. The fix for each is different.
Query-level analysis is how you find out which searches you actually lost after a Google update. In this guide we show you how to segment your Search Console queries by brand, intent and position band, how to tell ranking losses apart from click-through losses, and how to turn the results into a short list of actions. If you have not yet worked out which pages or sections were hit, start with page-level vs site-wide analysis, then come back here.
Key Takeaways
- Always split branded from non-branded queries; brand demand rarely depends on an algorithm update.
- Group queries by intent (informational, commercial, transactional, navigational) to see what kind of search the update changed.
- Use position bands (1–3, 4–10, 11–20, 21+) to separate real ranking losses from small shuffles.
- Compare impressions, clicks, CTR and average position together; a CTR drop with stable rankings is a different problem.
- Google’s guidance says small moves, such as position 2 to 4, do not need drastic action.
Set Up a Fair Comparison
Query analysis is only as good as the periods you compare. Google’s core update documentation recommends waiting until the rollout is complete, then at least a full week, and comparing that week with the week before the rollout began. Check the dates on the Google Search Status Dashboard first. For example, the dashboard lists the May 2026 core update as starting on 21 May 2026, with completion confirmed on 2 June 2026.
Google’s traffic-drop debugging guide also suggests looking at a 16-month view to check for annual patterns, and comparing year over year. A query that drops every June is seasonality, not an update.
Step-by-Step Query Segmentation
- Export queries for both periods. In the Performance report, use Compare mode and export the Queries table. If you have large data, the Search Console API or a Looker Studio connector avoids row limits in the interface.
- Split branded and non-branded. Google added a branded queries filter to the Performance report (announced in November 2025 and expanded to all eligible sites in March 2026, according to Search Engine Land). It uses an AI-assisted classifier, so spot-check it. Sites that are not eligible can use a regex on brand names and misspellings; see our Search Console regex guide.
- Tag intent. Use simple rules first (words like “how”, “what”, “why” for informational; “best”, “vs”, “review” for commercial; “buy”, “price”, “near me” for transactional), then hand-check the top queries.
- Assign position bands for both periods based on average position.
- Calculate the change in clicks, impressions, CTR and position for each query and each segment.
- Rank by lost clicks, not percentage, so small queries with big percentage swings do not distract you.
- Map queries to landing pages to see whether the same page lost many queries or many pages lost one kind of query.
Reading Branded vs Non-Branded Changes
Branded queries are searches for your company or product names. They mostly reflect awareness and demand, so they tend to be stable across algorithm updates. If branded clicks fell in step with non-branded clicks, look for causes outside the algorithm: a seasonal slump, a PR issue, a site outage or a tracking change. If branded traffic is stable and non-branded fell, the update is a more plausible explanation. Our guide to branded vs non-branded keywords covers the strategy side, and using the branded queries filter for update analysis covers the tool itself.
Position Bands: What Kind of Loss Is It?
| Movement | Example | How to read it | Typical response |
|---|---|---|---|
| Within top 3 | Position 2 → 4 | Small shuffle; Google says no drastic action is needed | Monitor; avoid rewriting a page that still performs |
| Top 3 → page one lower half | Position 3 → 8 | Real loss of visibility on that query | Compare your page with the pages that now outrank you |
| Page one → page two | Position 7 → 15 | Significant demotion for that query | Check relevance, depth and intent match |
| Large fall | Position 5 → 40+ | Page may no longer be seen as a good match | Review whether the page should target this query at all |
| Impressions gone | Query no longer appears | Page dropped out or the query was re-mapped to another URL | Check for cannibalization and indexing |
Remember that Search Console’s average position is the average of the topmost result from your site, so one URL being replaced by another of your URLs can blur the picture. Our guide to tracking rankings accurately explains where rank trackers and Search Console disagree.
Rankings vs Clicks: Separate the Two Problems
Google’s debugging guide makes a helpful distinction: if impressions hold steady but clicks fall, look at your titles and snippets rather than assuming a ranking problem. In 2026 there is a further layer, because SERP features such as AI Overviews can reduce clicks on queries where your position did not change.
| Impressions | Average position | Clicks / CTR | Most likely explanation |
|---|---|---|---|
| Down | Down | Down | Ranking loss; continue with page and SERP analysis |
| Stable | Stable | Down | SERP layout change, AI Overview or snippet issue |
| Down | Stable | Down | Lower demand or seasonality; check Google Trends |
| Up | Down | Stable | Ranking for more, broader queries at lower positions |
We break down the click-side scenario in AI Overviews CTR vs ranking loss and the wider difference in rankings vs traffic after an update.
Intent Segments: What Did the Update Change?
Once queries are tagged by intent, look for a lopsided result. Some typical findings:
- Informational queries fell, commercial held: other pages may answer these questions more directly, or the SERP now favours forums, videos or AI summaries.
- Commercial “best” and “review” queries fell: look at how much first-hand testing and original evidence your pages show compared with the new top results.
- Local or “near me” queries fell: check whether the drop is in the organic results or in the map pack, which is affected by different factors.
- One topic fell across intents: this points to topical relevance; study the winners with a SERP competitor analysis.
If the dominant intent of a query changed (for example, a SERP that used to show guides now shows product listings), read our explainer on search intent before you rewrite anything.
Common Mistakes
- Mixing branded and non-branded queries, which hides the real non-branded loss.
- Sorting by percentage change instead of absolute clicks lost.
- Comparing a mid-rollout period with a pre-update period.
- Treating every CTR drop as a ranking drop.
- Rewriting pages for queries where you moved only one or two positions.
- Ignoring anonymised or very rare queries: Search Console does not show every query, so totals in the table may not match the chart.
“There’s no need to take drastic action”
— Google Search Central, on small ranking drops such as position 2 to 4, Google Search's core updates and your website
Related Guides
- Who Took Your Rankings? SERP Competitor Analysis After an Update
- Running a Content Quality Audit After a Core Update
- Delete or Improve? Handling Weak Content After an Update
Frequently Asked Questions
What is query-level analysis?
It is the process of comparing individual search queries before and after an update to see which searches lost impressions, positions or clicks, and grouping them by brand, intent and position band.
Why separate branded queries?
Branded searches mostly reflect demand for your company, which algorithm updates rarely change. Mixing them in can hide or exaggerate a non-branded loss.
What date range should I compare?
Wait until the update finishes rolling out, then compare at least one full week after completion with the week before the rollout started. Check longer windows and year-over-year too.
My rankings are stable but clicks fell. Was I hit by the update?
Probably not in the ranking sense. Stable impressions and positions with lower clicks usually point to SERP layout changes, AI Overviews or snippet issues.
Should I fix every query that dropped?
No. Focus on queries with meaningful lost clicks and large position changes. Google advises against drastic action for small moves.
Conclusion
Query-level analysis turns a scary traffic graph into a clear list: which searches you lost, how badly, and why it might have happened. That list is what makes the next steps targeted and calm. If you would like help building the segmentation and turning it into an action plan, our SEO team can do the analysis with you. Get a quote to start.
References
- Google Search's core updates and your website (Google Search Central)
- Debug Google Search traffic drops (Google Search Central)
- Performance report (Search results) – Search Console Help
- Google Search Console branded queries filter expands (Search Engine Land)
- Google Search Status Dashboard: ranking incident history
