Article Ideas from Search Console Queries: Impressions Guide

The Queries tab in Search Console lists the searches that made your site appear in Google results, which makes it the most direct raw material you have for new article ideas. A query with many impressions and few clicks marks a place where Google showed your page and the searcher either did not see enough reason to click or wanted something the page does not provide. If you run a blog, a publication or a content team, the job is to filter that list, group the survivors by intent and decide for each group whether it deserves a new article, a new section in an existing one, or a pause. The short rule: a query becomes a new article only when no existing page answers the intent behind it.

How to read the report itself is covered in Search Console performance report analysis. Here the focus is what comes after reading: the path from a query to a topic to a brief. If you use the Search Console interface in another language, menu labels may differ from the English labels used here.

The query list as a topic source

The Performance report help page (Opens in a new tab) says the queries dimension groups data by the search query users typed. The list is therefore a record of real search behaviour, not an estimate. Every row is a search for which your site appeared in results at least once.

The SEO Starter Guide (Opens in a new tab) points out that people who know a topic well may use different words than newcomers, and that Google's language matching systems are sophisticated enough to relate pages to queries without exact word overlap. Together these explain why the report is useful: it shows how searchers describe a subject, which can differ from the labels your own team uses, and it shows that your pages already appear for some of that language.

The report has a built-in limit. It lists only searches where your site was shown. Topics you have never covered produce no impressions, so they never appear. For those, keyword research remains a separate input. The Queries tab finds opportunities around your existing visibility; it cannot find gaps from nothing.

Preparing the data before you read it

According to the help page, the default view covers the last three months, the date filter changes the range, and the search type can be web (text-based), web (multimodal), image, video or news. Before looking for ideas, prepare the view.

1. Set the search type to web so image and video intents do not mix into article ideas.
2. Choose a date range long enough to avoid a single news spike, and compare it with an earlier period for seasonal subjects.
3. Check the Countries tab to confirm that the queries come from the audience the article will serve.
4. Check the Devices tab to see whether mobile and desktop searchers ask differently.

The page also says the newest data may be preliminary and can change within a few hours, shown as a dotted line on the chart. Do not decide on a topic from the last day or two of data.

Moving the data into a working sheet

Practical suggestion: copy the rows you keep into your own sheet and add columns for intent label, related existing page, decision, decision date and notes. The decision date matters. When the same query returns three months later, you can see what you decided before and why.

One more preparation habit saves time later: write down the date range, search type and country you used next to the list. When a colleague opens the sheet next month, they should be able to reproduce the same view and see whether the groups moved. Without that note, two people comparing the same query can reach different conclusions simply because their filters differed, and the discussion turns into an argument about data instead of a decision about content.

Finding queries with many impressions and few clicks

The raw list can run to hundreds of rows. The point of filtering is to reach a handful of queries worth planning around. Practical suggestion for the order of work:

1. Sort by impressions and start from the top.
2. Remove branded queries, those containing your site, author or series names. They are navigation, not ideas.
3. Flag rows with a low click-through rate.
4. Among the flagged rows, keep those whose average position lags behind the rest of your site.
5. Write the survivors on one page.

Documentation sets no cut-off for any of this, and every site has a different traffic level, so judge against your own averages. A row with clearly higher impressions and lower clicks than your other queries is a candidate.

The help page defines the metrics: clicks as the number of times a user clicked through to your site from Google Search results, impressions as how many times your site appeared in results, click-through rate as clicks divided by impressions, and average position. One detail changes how you read position: in the chart it reflects the topmost result across the whole site, while in the table it reflects the specific URL or grouping in that row. Use the table rows for topic decisions. Also note that two results from the same site for one query count as one impression.

Three reasons a query underperforms

Only the second and third reasons can lead to new content. The first is a presentation problem on a page that already works. The SEO Starter Guide describes a good title as unique to the page, clear and concise, and an accurate description of its contents, and it says you have complete control over the words that can be used to generate a snippet. Fixing presentation should come before writing anything new.

A five minute presentation test

Type the query yourself and read your result. Does the title and description tell a searcher that this page answers their exact question? Compare it with competing results: which piece of information do they show that yours does not, such as a year, a section name, a date or a comparison cue? Adjust without exaggeration. A headline that overpromises may raise clicks briefly and still disappoint the reader.

Grouping queries by intent

Different spellings of one question are one idea. Queries such as "when is the festival programme announced", "festival programme release date" and "next year festival dates" describe one need. Give each group one topic card and tag it with an intent label.

Choosing a format from intent is its own subject. At this stage the label only prepares the group for a brief.

Separating navigation from discovery

Branded and navigational queries show that readers already know you. They distort the click rate picture because readers who type your name already know you, and they hide the discovery queries that matter for new ideas. Park them in a separate list and look at them only when a branded query shows impressions without clicks, which is a different problem.

Decision rule: new article or new section

For each group, find which of your pages currently appears for it. Open the query and switch to the pages view to see. Then answer three questions.

The third question is the cannibalization check. When two pages target the same intent, it becomes unclear which one should be shown; see how to detect keyword cannibalization. On a site with many similar pages, the content inventory decision matrix helps before you commit.

Decisions by situation

When one query hides several intents

Some queries cannot be placed under one label. A short phrase about a festival may come from a reader who wants the latest news, a reader who wants the screening schedule and a reader who wants the history of the festival. Your page appears for all of them, clicks are low and the average hides the split. Test by reading the live results page: if the top results are a mix of formats, the intent is mixed. In that case do not write one article for the whole phrase. Split the group into sub-intents, assign each to the page that serves it best and give the weakest sub-intent its own section or article.

Turning a query group into a brief

Once a group is confirmed as a new article, it fills several brief fields. For the full method, see how to write an SEO content brief.

• Reader question: the clearest query in the group, in the searcher's own words.
• Related phrasing: the other queries, as natural language for headings and sections.
• Link plan: which existing page links to the new one and which page the new one links back to.
• Addition: the section that satisfies the need in a way no existing page does.
• Follow-up plan: which queries will be watched after publication.

The Starter Guide advises creating content from what you know about the topic and not rehashing what others have published. The query list names the topic. The originality has to come from your own observation, testing or reporting.

Worked example: a film festival news site

Example scenario: A film festival news site covers an annual festival in one city. Its editor opens the query list limited to web search, removes branded queries and flags rows with strong impressions and weak clicks. Three groups stand out.

The first group concerns a jury rule: how the festival decides when two films receive equal votes in a competition section. The site's coverage mentions the rule in a sentence, and those pages appear for the queries. The intent is "learn", and the existing pages answer only a corner of it. Decision: add a short explainer section to the festival guide page instead of opening a new article.

The second group concerns how accreditation works for freelance critics at smaller festivals. No page fits the intent. The newsroom has a reporter who covers regional festivals and can describe how those festivals handle accreditation in practice. Decision: new explainer, with the reporter's first hand notes as the addition.

The third group asks which of two ticketing options to use for catching screenings. The site has a page on screening schedules but no comparison. Because schedules change, the editor decides the comparison page must carry a visible review date and an owner. Decision: new decision guide, with a scheduled review.

The same list produced three decisions: add a section, write a new article and write a new article with a review plan. The value of the report is not that it produces articles but that it makes the right call for each row.

Following up after publication

Mark the group as "watching" in the sheet when the article goes live. After a few weeks, reopen the same queries and ask whether impressions rose, whether clicks followed and which page is showing.

These readings are prompts, not verdicts. The documentation sets no waiting period after publication, so change one thing at a time and note the date. For pages that lose ground later, see content decay.

Reporting the outcome to editors or clients

Query lists are noisy for anyone outside the process. Report decisions, not rows. A useful monthly note has four short parts: which groups were reviewed, which decision each received, which articles are now being watched and what changed in the groups watched last month. For a client, add one line per decision explaining why in plain language, such as "the existing guide answered only part of the question, so a section was added". This keeps the report honest about the fact that not every opportunity becomes a new page.

What a decision log looks like

One row per group is enough: the date, the group name, the intent label, the related page, the decision, the reason and the review date. After a few months the log shows which kinds of decisions paid off, and which kinds were guesses that need a stricter test.

Limits of the data

The Performance report help page does not state row limits, how anonymized queries are handled or how long data is retained. Check the interface notes and the current help pages for those details. Limits in Google's documentation can change; check the current page before relying on any figure.

What an impression actually counts

An impression means your site appeared. It does not measure total search volume. When your position lags, impressions can understate real demand, and when it leads they can overstate your reach for a topic you only partially cover.

Why the average hides the spread

Average position blends different rankings shown to different people at different times. Use it to see whether a query lags behind your other queries, not to quote an exact rank.

Thin rows and noisy rates

For queries with few impressions, one or two clicks swing the click-through rate. Use such rows for ideas, but look for more evidence before acting.

Festival calendars

Programme announcements, submission deadlines and the festival week itself move query volume. A three month window can capture one event and miss another. Practical suggestion: compare a key group over two periods and check the calendar before reading a spike as lasting demand.

A monthly routine

A thirty minute monthly meeting with four agenda items keeps the sheet honest: groups reviewed, decisions made, articles being watched and ideas rejected.

Why record the rejections

A rejected idea is not wasted. Write the reason beside it. When a new test, interview or dataset arrives, the idea can return with an addition it lacked before.

Checklist

• Search type is web and the country matches the audience.
• Branded queries are out of the idea list.
• The last day or two of preliminary data did not drive any decision.
• Each candidate group has an intent label.
• Each group was compared with existing pages.
• The cannibalization check was done before choosing a new article.
• Each confirmed group has a brief with an addition.
• Queries to watch after publication are recorded.

Mistakes to avoid

Judging a query by its row alone

A single row hides which page appeared and what the searcher found there. Open the query, check the pages view and read the live result before assigning a decision.

Reading the list once and never returning

A query group that was rejected in spring may deserve another look after a new observation or dataset arrives. Keep the sheet alive and date every decision.

Merging unlike intents into one group

Grouping by shared words puts a learner and a buyer under one card. Group by what the searcher wants to achieve, and split the group when the live results disagree.

Treating a title edit as a cure for thin content

A sharper title can lift the click rate, but if the page answers only part of the question, the reader still leaves. Decide first whether the problem is presentation or coverage.

Leaving the working sheet without a decision date

Without dates, the same query returns three months later and nobody remembers why it was parked. One date column prevents repeated arguments.

Frequently Asked Questions

Can the query report alone tell me what to write next?

On its own it is not. The report holds only searches where your site was shown, so topics with zero impressions never appear. Combine it with reader questions and keyword research.

A query shows plenty of impressions and few clicks. What now?

First compare the existing page title and description with the query. If the page answers only part of the intent, add a section. If no page answers the intent, treat it as a new article candidate.

How long a window should I read?

The help page shows the last three months by default and lets you change the range with the date filter. For seasonal topics compare two periods, and avoid deciding from the latest preliminary days.

Is there a position that signals a new article?

The documentation sets no threshold. Practical suggestion: look at rows with high impressions whose position lags behind your other queries.

Is it a problem to answer one query on two pages?

If both target the same intent, it becomes unclear which page should appear. Check existing pages first and strengthen the current page where that fits.

How much history does the report keep?

The Performance report help page does not state a retention period. Check the date options in the interface and the current help page.

Conclusion

The Queries tab turns idea generation from guessing into reading a record. Filter for high impressions and weak clicks, group by intent and choose one of the outcomes for each group: rewrite the presentation, add a section, write a new article or wait. Metric definitions come from the Search Console help page, and the principles about user language, original content and titles come from the SEO Starter Guide. Thresholds and the monthly routine are settings you adapt to your own data.

This month, do one thing: filter your query list to web search, pick five candidate rows and fill in the decision table for each. If you want a team to take over writing the articles that come out of that list, see our content writing service.