How to Find Blog Topics That Rank (a Keyword-First System)

Most blog calendars are built backwards. A team sits in a room (or a doc, or a chat window) and asks “what should we write about?”, generates twenty ideas that sound interesting, and ships them over a quarter. Six months later, three posts get traffic and seventeen don’t, because nobody checked whether anyone was searching for any of it.

The fix is to flip the order: start from demand, then get creative. Here’s the keyword-first system, six steps you can run manually in an afternoon, or automatically with the right tooling.

Step 1: Mine Search Console for queries you almost rank for

Your own Search Console data is the most underused topic source in SEO. Open the Performance report and look for queries where you sit in positions 8–30: real impressions, few clicks. Google already believes you’re relevant to these; it just doesn’t have a page from you that deserves page one. Each cluster of these queries is either a refresh (if an existing page is close) or a new post (if nothing on your site really answers it). These are your fastest wins, demand you’re one good post away from capturing.

Step 2: Expand your seed keywords

Take your core themes (the 5 to 10 phrases that define your business) and expand each into the long tail with a keyword tool: questions, comparisons, “how to” variants, “best X for Y” modifiers. Pull search volume and keyword difficultyfor everything. You’re not committing to any of it yet; you’re building the raw pool that the next steps will filter hard.

Step 3: Map competitor gaps

List the 3 to 5 sites that compete with you in search (which may not be your business competitors), and find the queries they rank for that you don’t. Gaps where multiple competitors rank are validated demand in your exact niche. Gaps where only one giant ranks are usually authority plays; note the adjacent, more specific versions of those queries instead.

Step 4: Classify intent and match the format

Label every candidate by intent: informational (learn), commercial (compare), transactional (buy). A blog post is the right weapon for informational and most commercial-investigation queries. Transactional queries belong to product and service pages, sending a blog post after them wastes the effort. Intent also dictates format: listicle, how-to, comparison, or definition, because Google rewards the format searchers expect.

Step 5: Validate the SERP before you write

For every surviving idea, look at what currently ranks. Three questions:

• Do blog posts actually win this query?If page one is all product grids, tools, or videos, a post can’t win regardless of quality.
• Can you realistically compete?A SERP of niche blogs and forum threads is winnable; a wall of household-name domains usually isn’t (yet).
• Can you add something the SERP lacks?Fresher data, real experience, a sharper angle, a better format. If you’d just be restating the top result, pick a different fight.

Step 6: Score, dedupe, and queue

Roll volume, difficulty, intent, and business relevance into a simple score so ideas can be ranked instead of debated. Then dedupe against everything you’ve already published: if a new idea overlaps an existing post, the right move is usually a refresh, not a new page that competes with your own URL. What survives is your queue: write from the top.

From topic to brief

A topic alone still under-specifies the post. Before writing, fix the primary keyword (the query the post owns), 2–4 secondary keywords (subtopics and variants for headings), the angle (what this post says that the SERP doesn’t), and the format the SERP rewards. That one-paragraph brief is what keeps the draft, human or AI, pointed at the demand you validated.

The takeaway

Finding blog topics isn’t a creativity problem; it’s a research problem with a creative step at the end. Mine what you almost rank for, expand from real keyword data, steal validated gaps from competitors, respect intent, check the SERP, and score what survives. Creativity then gets spent where it pays: on the angle and the writing, not on guessing what people might search for.