How to select and analyze search queries: define your goals, then identify queries worth improving

When selecting search queries, first decide which outcomes you want to improve, then define the scope of your data. Not every term you rank for should become an optimization target. If your goal is to improve your website’s organic search performance, start with queries already appearing in Search Console that are relevant to your content and users’ needs. Consider average position, click-through rate, and clicks together to identify opportunities worth pursuing. This guide focuses on search performance analysis for existing websites; these methods should not be treated as keyword selection rules for advertising.

Define your goals and data scope first

Whether a low click-through rate needs attention depends on your website’s goals. For example, users searching for opening hours, an address, or a phone number may already find that information in the search results. If your goal is to attract customers to a physical location, you may not need to turn every search into a website click. If you want users to visit your site to read more, check whether your titles and descriptions clearly explain why they should click through.

Google’s bubble chart example uses Search Console’s “Site Impression” data, aggregated by site and query. Before analyzing it, select:

• The website property and date range to analyze.
• Which queries and countries or regions to include or exclude.
• Which device categories to examine, and whether to compare mobile and desktop.

Select relevant queries before assessing opportunities for improvement

A low ranking does not mean a query has no value. For queries important to your website, Google’s article recommends prioritizing those already appearing in search results over those with no impressions at all. However, you should still confirm their relevance to your content rather than simply chasing rankings.

At Procrustes, we turn the questions our clients genuinely care about into articles, with ongoing content planning, writing, and updates.

• Low ranking, high click-through rate: These queries may closely match users’ needs. Check whether your existing content answers their questions adequately. For example, if “how to build a wooden cabin” appears among a gardening website’s queries, you could create a page on the topic if none exists, or fill in missing information on an existing page.
• Low ranking, low click-through rate: First distinguish relevant queries from irrelevant ones, then examine which still generate more clicks. Do not dismiss the entire group.
• Non-target queries: If they do not affect your goals, you can leave them alone. If you do not want this traffic, you can try adjusting the wording that causes your site to appear for these queries.

Use bubble charts to find clues, not jump to conclusions

A bubble chart can show click-through rate on the horizontal axis and average position on the vertical axis, with bubble size representing clicks and color distinguishing devices. Reversing the position axis so that position one appears at the top makes the chart easier to read. A logarithmic scale can help reveal extreme values, while mean or median reference lines can highlight queries that deviate from the overall pattern. Using colors to distinguish too many categories can make differences harder to see.

High and low values in the chart are relative to the selected data and thresholds, not universal benchmarks for all websites. A high ranking with a low click-through rate may mean other search results are more appealing, or that users have already found their answers. Review the query, relevant pages, and how the search results are presented. Do not assume a cause based solely on which quadrant a query falls into.