Google Search Console Blog Topics: How to Find What to Write

The thing that stops most blogs isn't writing time. It's the question "what do I even write about?" And the answer is already sitting in your own Google Search Console account — the exact words people typed before your site showed up. Not an estimate, not third-party search volume, but real queries on your domain.

Here's how I turn that data into blog topics, so every article has a reason to exist before a single word gets written.

What Search Console gives you — and what it never will

So you're not working with false expectations:

• Data goes back 16 months. Anything older, you have to archive yourself.
• Reporting lags roughly 1 to 2 days. Yesterday is usually incomplete.
• Some queries are anonymised. Google hides rare queries that could identify an individual user. Your query totals will never match total clicks — that's normal, not a bug.
• The interface export caps at 1,000 rows per table. Through the Search Console API you get far more, thousands of rows per request.
• You won't find topics your site has never appeared for. Search Console only shows you where you already leave some trace.

That last point matters. GSC is excellent for expanding what already half-works. It's useless for discovering a category you've never touched.

Step 1: Set the report up so it shows something useful

Open Performance → Search results, then:

1. Turn on all four metrics: clicks, impressions, CTR and average position. Without impressions and position you can't pick a topic.
2. Set the range to last 12 months, not 3 months. A short window flattens seasonal queries.
3. If you only sell in one country, add a Country filter. Otherwise random foreign impressions leak into the list.
4. Switch to the Queries tab and sort by impressions, not clicks.

Sorting by impressions is the whole trick. Clicks show you what already works. Impressions show you what people wanted and didn't get a good enough answer for from you.

Step 2: Four patterns that turn into topics

In the query table I look for four situations. Each one calls for a different fix.

1. High impressions, position 8 to 20, few clicks

The best source of topics there is. Google already understands your site is related to the query, but you have no page that covers it directly. Usually what's ranking is a category page, the homepage, or an article about something adjacent.

The fix is a dedicated article on exactly that query. Moving from position 14 to position 5 means several times more clicks at the same impression count.

2. Good position, terrible CTR

A query sitting at position 3 to 5 with CTR under one percent is not a new-article topic. It's a title and description problem. People see you in the results and scroll past.

New content won't help here. Rewriting the title tag and meta description of the existing page will. The same logic applies to ecommerce category pages, where titles tend to repeat across dozens of URLs.

3. Question queries

Filter for queries containing "how", "why", "what is", "when", "how much", "which". Type one word into the filter box and look at what comes back. These are almost always informational, and almost always belong in an article rather than on a product page.

Bonus: a question from GSC is a finished headline. You don't have to guess how people phrase things — you have their exact wording.

4. The wrong page is ranking

Click a query, then switch to the Pages tab. If two or three of your pages take turns appearing for one query, they're competing and neither gets full weight. Instead of writing another article, merge the content into one strong page.

Step 3: Pair every query with a page

A query list on its own isn't enough. You need to know which page currently shows for the query, otherwise you can't tell whether to write something new or extend what exists.

How: Pages tab → click a specific URL → switch to the Queries tab. You'll see every query pulling that page into results. Very often one article is collecting impressions across a dozen related topics — and some of them deserve their own text.

Step 4: Group queries into topics

One query is not a topic. A topic is a group of queries with the same intent.

Export to a spreadsheet and group rows by meaning, not by exact wording. "How to fix duplicate title tags", "duplicate title tags" and "same title on multiple pages" are three queries and one article.

A working rule: a cluster with at least 5 to 15 related queries and a few hundred impressions a year between them can carry a standalone article. Smaller clusters go in as a section of something you already published.

Step 5: Prioritise by realistic gain

Rank clusters by a simple estimate: cluster impressions × chance of moving up. The chance is highest for queries at positions 8 to 20. Above position 50, more than one article stands between you and the first page — no point pretending otherwise.

Write the first three to five topics and park the rest. Producing twenty articles at once makes no sense. Positions move in weeks, not days, and without measurement you won't know which of them did anything.

Step 6: After publishing, check whether it worked

A data-driven topic without measurement is still just a better guess. Once the article is live, set up a 28 days before / 28 days after comparison in Search Console and watch three things on that specific URL:

• the number of queries the article shows up for (this almost always moves first),
• average position on the target queries,
• clicks.

Give it 4 to 8 weeks. A new article has to be indexed and classified first. Anyone claiming you'll see the effect in a few days is describing an exception, not the rule.

Where this process breaks down

You can do all of this by hand on a site with a few hundred queries. On an ecommerce store with dozens of categories and thousands of queries you hit two walls: the 1,000-row export limit and the hours grouping takes. That's why most blogs stall after two months — not on the writing, but on the deciding that comes before it.

How I do it

You connect Search Console and I pull queries through the API, so I'm not capped at a thousand rows from the interface. I group queries into topics by intent and separate the ones that need a new article from the ones where rewriting a title or meta description on an existing page is enough. For each topic I propose a specific headline and write the finished article — not an outline, not a note saying "you should cover this".

After publishing I measure the same 28-days-before-and-after window, so for every article you know what it brought in. Including when the honest answer is: nothing yet.

Quick checklist

1. Performance → Search results, 12 months, all four metrics, country filter on.
2. Queries tab, sort by impressions.
3. Mark queries at position 8 to 20 with high impressions.
4. Split off queries with good position and low CTR — that's title work, not article work.
5. Filter for question words.
6. Pair queries with pages, find competing URLs.
7. Group into topics, prioritise the first three.
8. Publish, compare after 28 days.

#seoobsah