High impressions, low clicks: where your next content idea may be hiding

A content idea does not need to start in a keyword tool. Sometimes it is already in the report you open every week.

The point is to find queries where the site shows up but does not yet deliver a good answer to the person searching.

The filter

In Search Console, go to Performance → Search results → Queries.

I would start by looking for three signals at the same time:

• meaningful impressions: there is enough demand for the query to show up often;
• lower CTR than expected: the page is being shown, but few people choose to click;
• average position roughly between 8 and 20: the result is already in play, but still has room to gain visibility.

These numbers are a starting point, not a rule. The range changes depending on the site, the type of query, the intent and the page that is ranking.

What this filter actually tells you

It does not prove that “the content is bad.”

There may be a title and snippet problem. There may be an intent different from the one the page serves. It may be a very competitive query. It may simply be a SERP where the user gets the answer without clicking.

The filter is useful for something else: it creates a list of hypotheses worth investigating.

What I do with the list

The only step where I would bring AI into this workflow is grouping.

I paste the queries and ask the model to:

1. group variations that seem to share the same intent;
2. name the intent of each group;
3. not suggest titles;
4. not write content.

The goal is to cut the noise before deciding what deserves a new page and what should improve a page that already exists.

The mistake I have made

Treating each query as an article.

This can produce several thin pages competing for the same intent. Instead of increasing coverage, you increase the amount of content you have to maintain.

Today I prefer a simpler question:

Does this query deserve a new page, or a better answer where it already exists?

It is a small difference in the process and a huge one in the amount of unnecessary content you produce.

The limit

Search Console shows how Google and users behave on your site. On its own, it does not explain why a page lost a click.

That is why I use the filter as a starting point for investigation, not as an automatic diagnosis.

The important part is still interpreting the intent and looking at the page that is getting the impression.