FIELD GUIDE · REVIEWED OCTOBER 15, 2026

Find query and page opportunities

Use query and page performance to guide investigation. Search Console shows returned searches and landing pages with clicks and impressions, but the rows are not a complete census.

Start with a measurable question

A useful opportunity begins with a query worth inspecting. This tool returns query/device candidates; use get_search_analytics grouped by query and page to investigate their landing pages separately. For example, a query may have many impressions but a modest click-through rate, or a relevant page may appear around positions 4–20. These are clues, not diagnoses: intent, result features, seasonality, and query mix affect performance. Search Analytics reports aggregated metrics for the dimensions you request. Google may omit anonymized queries and other rows, so the returned data should be treated as partial evidence rather than a complete inventory.

How the directional estimate works

This report groups returned query rows by device and compares each eligible query with other returned queries on the same device and rounded average position. A candidate must rank from position 4 through 20 and meet the impression threshold, which defaults to 50. Its peer group needs at least five other queries and an impression floor of the greater of 1,000 or five times the candidate’s minimum-impression threshold. The reference click-through rate is total peer clicks divided by total peer impressions. The displayed upside is the positive difference between that rate and the candidate rate, multiplied by candidate impressions and rounded. Example: 20 clicks from 1,000 impressions is 2%. Peers with 100 clicks from 2,500 impressions yield 4%. The 2 percentage-point difference corresponds to 20 directional clicks (0.02 × 1,000), not a forecast or promise.

Turn a candidate into a test

Open the query and landing page, check whether the page directly answers the searcher’s likely need, and inspect the current result for your target location and device. Compare the title and snippet with the page content and competing results. If the page is a good fit, make one focused change and record what changed. Revisit the same query, page, device, and a comparable date window after enough time for results to settle. Keep the original metrics so you can distinguish an observed change from an assumption about why it happened.

Try the workflow

Explore the interactive sample first. For an approved connected account, ask your authorized MCP client:

For sc-domain:example.com, use find_striking_distance_opportunities for August 15–28, 2026. Explain each query/device candidate and its peer cohort. Then use get_search_analytics grouped by query and page for the same dates to investigate landing pages. Label directional estimates and coverage limits.

Relevant tool: find_striking_distance_opportunities. See product documentation and connection settings.

What the evidence cannot tell you

The calculation uses returned Search Analytics rows, which may omit anonymized or low-volume searches and may be truncated by row limits. Peer queries are not necessarily equivalent in intent or search-result layout. Rounded position and a site-specific peer CTR do not control for those differences. The estimate describes a mathematical gap against the selected peer cohort; it does not predict that a page change will earn additional clicks or establish a cause.

Sources and build scope

Original editorial examples, reviewed October 15, 2026. Google defines clicks, impressions and average position; its Search Analytics API reference documents grouping, inclusive Pacific Time dates, fractional CTR and bounded results.

Product behavior checked against Searchlight source 4f160c3, calculation version searchlight-insights-v2 (repository access required). This is implementation provenance, not proof of production availability or live-provider acceptance. Roadmap features are not described as shipped.

← All guides