Use matched reporting windows
Choose the exact Search Console property and two fully reported 28-day windows. Keep the same search type, country, device and query/page filters. Calendar months can have different lengths.
Search Console · SEO diagnosis
Compare Google Search Console performance between two periods
Compare clicks, impressions, CTR and average position across two complete windows using identical filters. Start with the site-wide change, then investigate the pages and queries behind it. A percentage change without the original numbers rarely explains what happened.
14 days free · No credit card
Google Search ConsoleIllustrationOrganic clicks · illustrative periods1,200 → 1,080
−120 clicks (−10%) on equal 28-day samples. Next: check impressions, query mix and position.
Previous 28 daysLatest complete 28 days
01 / Understand the signal
A fair comparison is more useful than a dramatic percentage
A decline of 10% can come from ranking changes, fewer searches or a different composition of queries. Compare exactly the same property, search type and filters, and avoid incomplete reporting dates. Show absolute and relative deltas together. CTR should change in percentage points; position improvements appear as negative position deltas. Neither metric explains causality on its own.
02 / The analysis01 02 03
Compare SEO periods without mixing unlike samples
Read all four metrics together
Report clicks and impressions as absolute and relative change; report CTR as a difference in percentage points, and position as an average difference. Never interpret unreturned query rows as zero.
Segment the real movement
Break down the biggest changes by page and query, then country or device if relevant. Check which pages account for the absolute click change, and flag anything the returned rows cannot explain.
03 / Worked example
Same −10% clicks, different reasons to investigate
Fictional matched 28-day example. Impression volume and CTR move in different directions, so an overall click delta cannot stand alone.
| Metric | Period A | Period B | How to read it |
|---|---|---|---|
| Clicks | 1,200 | 1,080 | −120 clicks (−10%): traffic outcome |
| Impressions | 30,000 | 36,000 | +6,000 (+20%): visibility increased |
| CTR | 4.0% | 3.0% | −1 percentage point: fewer clicks per impression |
All URLs, queries and numbers in this table are fictional. They demonstrate the method, not actual performance.
Read the evidence
Traffic fell even as impressions rose
In this fictional sample, impressions grew 20% while clicks dropped 10%. CTR moved from 4% to 3%: that's −1 percentage point, not −1%. A page or query breakdown could show expansion into lower-ranking searches, a genuine CTR issue or changes in search intent. It does not establish which cause applies without deeper checks.
04 / Ready-to-copy prompts
Three prompts to compare periods properly
AskGSC includes a period-comparison tool. The assistant can interpret its returned rows, but cannot reconstruct suppressed queries or external SERP changes.
AskGSC offers read-only Google Search Console access through MCP: accessible properties, Search Analytics and period comparisons. Use the exact authorized property, not a guessed domain.
Explore all 18 Google Search Console prompts01Compare the full property
Use gsc_compare_periods for [PROPERTY] on [PERIOD_A_START]–[PERIOD_A_END] and [PERIOD_B_START]–[PERIOD_B_END], two complete equal-length periods. Keep grouping and filters identical. Show clicks/impressions as absolute and relative deltas, CTR in percentage points and position in average positions. Never use zero for missing keys.
02Explain the page-level shift
For [PROPERTY], compare two complete equal 28-day windows grouped by page. Rank the 10 largest absolute click changes; show impressions, CTR and average position beside each. Investigate the returned queries on the leading URLs and distinguish data from hypotheses.
03Investigate a release date
For [PROPERTY], compare two complete equal-length GSC periods around [RELEASE_DATE], keeping the same country/device filters. Report site-wide changes and affected pages. A timing correlation does not prove the deployment caused the change; say what else must be checked.
05 / Decide what to do01 02 03
Three interpretations to test before acting
Clicks down, impressions stable
Inspect the query mix, CTR and positions. A snippet change is one hypothesis, not the default diagnosis.
Clicks and impressions both down
Check demand, seasonal patterns and rankings on the same queries before rewriting pages.
Impressions up, clicks flat
The site may be appearing for additional, less competitive queries. Group by query and page instead of assuming growth failed.
Equal dates do not guarantee identical search demand
Search Console reporting can lag, average position is aggregated, and query rows can be anonymized or omitted. AskGSC compares the returned samples with consistent dimensions and filters; it cannot guarantee complete keyword coverage, prove causality or read deployment logs.
06 / FAQ
Questions worth answering
What Search Console can and cannot tell you about this particular SEO problem.
01How do I compare two periods in Google Search Console?+
Use complete equal-length windows with identical property and filters, then compare clicks, impressions, CTR and position. Segment the affected pages and queries before interpreting causes.
02Should I compare week over week or month over month?+
Match the decision: 28-day windows help smooth weekday differences; year-over-year checks can help with seasonality. Keep period lengths and reporting completeness consistent.
03Is a negative average-position delta good?+
Usually, yes: a smaller position number is higher ranking. But weighted averages and a changing query mix can move without a comparable page-level improvement.
Run it on your own data
A useful SEO comparison ends with a next question.
Get comparable Search Console numbers into your AI conversation, then investigate the pages that account for the movement.
Compare my SEO periods14 days free · No credit card