Find and Test Keyword Opportunities in Search Console

Queries with impressions and average positions around the second results page can identify useful review candidates. They are not guaranteed quick wins. Investigate intent, the ranking page and the quality of the underlying data before changing an article.

Build a candidate list with context

Compare two similar date windows and note seasonality, device and country. Group related queries by the task they express. Check whether the same page serves those queries, whether another page already gives a better answer, and whether the page's visible content matches the intent. Average position aggregates observations; it does not mean every user saw a fixed position.

Prioritize an actionable gap

Avoid creating a new page for every query variation. Similar wording can represent the same task, while the same word can appear in different intents.

Worked example: an incomplete troubleshooting guide

Suppose a robots.txt guide receives impressions for queries about indexing despite a crawl block. Review whether the guide explains that robots.txt is a crawl control, distinct from noindex. Add that distinction and a verification procedure if missing. Keep the URL stable unless there is a separate reason to consolidate it. This is an editorial example, not a forecast of traffic gains.

Evaluate the change fairly

Record the change date, target query group and expected user benefit. After recrawling and enough observations, compare similar query, device and country segments. Check conversions or completed tasks where available. A ranking shift can have other causes; do not attribute every movement to your edit or claim a promised uplift.

Use the official Search Console Performance report guidance to understand the metric definitions and reporting limitations.