You've got one keyword — say "email deliverability" — and eleven browser tabs open, and by tab six you're just guessing which subtopics matter and which are filler. That's the actual problem with learning how to map a topic cluster from a seed keyword: not the concept, which is simple, but the execution, which turns into an afternoon of copy-pasting "People Also Ask" questions into a spreadsheet with no idea if you've covered the topic or just found ten ways to restate the same idea.
Here's the fix. Mapping a topic cluster from one seed keyword means expanding that single term into a hub page and a set of spoke articles, grouped by what someone is actually trying to do when they search — not by how many keywords contain the same word.
How Do You Map a Topic Cluster From a Seed Keyword?
To map a topic cluster from a seed keyword: pull 50-100 related terms using autocomplete, People Also Ask, and a keyword tool; group them by search intent rather than surface similarity; pick the broadest, highest-volume term as your pillar; assign every remaining group as a spoke; then link every spoke back to the pillar and to two or three sibling spokes.
That's the whole mechanic. The hard part is step two — grouping by intent — because it's the step everyone skips.
- Expand the seed into a raw keyword list. Type your seed into Google, screenshot the autocomplete suggestions, then scroll to "People Also Ask" and "related searches" at the bottom. Repeat with two or three phrasing variants of the same seed. You want 50-100 raw terms before you touch a spreadsheet.
- Pull search volume and difficulty for the full list. You don't need every term ranked, you need to see which ones cluster around meaningful volume versus which are long-tail noise with three searches a month.
- Group terms by what the searcher wants, not by shared words. "Email deliverability checker" and "why emails go to spam" both contain deliverability-adjacent language, but one wants a tool and the other wants an explanation. Different intent, different spoke.
- Name each group with a working title. If a group can't be summarized in six words, it's actually two groups pretending to be one.
- Pick your pillar. The pillar is the group with the broadest intent and the highest realistic search volume, usually the definitional or "ultimate guide" style query the seed keyword itself represents.
- Assign remaining groups as spokes. Each spoke should answer one specific sub-question the pillar can only mention in passing.
- Map the internal links before writing anything. Every spoke points to the pillar. Adjacent spokes point to each other. The pillar points down to every spoke. Draw this as an actual diagram, not a mental note, you will forget which spoke needs which link by article four.
- Sequence the writing order by commercial and search value, not alphabetically or by whatever's easiest to write first.
That sequence is the entire framework. Most guides stop there. The part that actually determines whether the cluster ranks is what happens in step three, and almost nobody explains it properly.
The Grouping Mistake That Breaks Most Clusters Before They're Published
Here's the part every other guide on this topic glosses over: people group keywords by shared vocabulary instead of shared intent, and it quietly kills the cluster before a single article is written.
Say your seed keyword is "topic cluster SEO." You'll pull dozens of variants — "what is a topic cluster," "topic cluster examples," "topic cluster template," "topic cluster vs silo," "how to build a content cluster." A founder in a hurry looks at that list, sees the word "topic cluster" repeated ten times, and assumes it's one big spoke. Wrong. "What is a topic cluster" is informational — someone new to the concept wants a definition. "Topic cluster template" is a bottom-of-funnel searcher who already understands the concept and wants a tool to execute it, today. Those are two different articles serving two different reader mindsets, and if you cram them into one page, you end up with a 3,000-word article that satisfies neither search intent fully, Google notices, and so does the reader who bounces at paragraph four.
The tell that you've grouped wrong: if you can't write a single, specific meta description for the group without using the word "and" three times, you've merged two intents. Split it. This single fix, grouping by intent rather than keyword overlap, is the difference between a cluster that ranks six spokes out of eight and one that ranks two.
Cannibalization is the other failure mode hiding in this step. If "topic cluster examples" and "topic cluster template" both promise to show you real cluster structures, you've built two pages competing for the same click, and Google will pick one and suppress the other. Before you finalize your spoke list, run a pass specifically checking for overlapping intent between adjacent groups — it's worth reading up on how to fix keyword cannibalization if this is new territory, because it's a mistake that's expensive to fix after fifteen articles are live and nearly free to fix before you've written one.
What a Topic Cluster Actually Looks Like Once It's Mapped
A mapped cluster isn't a mind map with bubbles, it's a table. One row per spoke, with columns for the target keyword, search intent, estimated volume, and the two internal links it needs to send and receive. If your map doesn't fit that structure, it's still an idea, not a plan.
Picture the "email deliverability" example fully mapped out. The pillar is "email deliverability guide," sitting at maybe 2,400 monthly searches. Underneath it: a spoke on "why emails go to spam" (troubleshooting intent), a spoke on "SPF DKIM DMARC explained" (technical setup intent), a spoke on "email deliverability checker tools" (commercial intent), and a spoke on "cold email deliverability" (a narrower audience with its own vocabulary). Four spokes, four distinct reasons someone opens a search bar, zero overlap. That's a cluster doing its job.
Compare that to a badly mapped version: the same seed produces "email deliverability tips," "improve email deliverability," "email deliverability best practices," and "how to increase email deliverability" as four separate spokes. Read those four titles back. They're not four intents, they're one intent, phrased four ways. Publishing all four either triggers cannibalization or, worse, you write four thin articles instead of one strong one. A hub-and-spoke structure only works when each spoke earns its place by answering a question the pillar can't fully answer on its own.
Where to Get the Raw Keywords Without Paying for Ahrefs or Semrush
You don't need an enterprise SEO subscription to build the raw list, you need three free sources and forty-five minutes. Google autocomplete, "People Also Ask," and Google Search Console's existing query data (if your site has any traffic at all) will get you 70% of the way there for free. The remaining gap — what your competitors rank for that you've never even considered as a spoke — is the part free tools genuinely can't show you, because autocomplete only reflects what people search, not what's already ranking and winning clicks.
That's the actual limitation worth naming plainly: manual research finds keywords people type into Google, but it doesn't show you which of those keywords a competitor already owns, at what position, with what content format. Skipping that check means you might spend three weeks writing a spoke for a term where a competitor's three-year-old article with forty backlinks already sits at position one, a fight you didn't need to pick. This is the exact gap keyword gap analysis tools exist to close, and it's worth running one pass through competitor keyword research before you lock your spoke list, specifically to catch terms your competitors rank for that never showed up in your autocomplete pull.
WriteGap's topic cluster planning feature does the grouping-by-intent step automatically once you drop in a seed keyword, mapping hub and spokes with the search volume and difficulty attached to each node, which turns the forty-five-minute manual sort into something closer to five. It also flags where a proposed spoke overlaps too closely with an existing page on your own site, catching cannibalization before it's a live problem instead of a six-month-old one.
How Many Spokes Should a Topic Cluster Have?
There's no fixed number, and anyone who tells you "always eight spokes" is guessing. The right count is however many distinct search intents actually exist under your seed keyword — for a narrow B2B topic that might be four; for something broad like "email marketing," it could legitimately be twenty or more, published over months rather than all at once.
What matters more than the count is sequencing. Publish spokes in order of commercial value and realistic ranking difficulty, not in the order they occurred to you. A spoke targeting a low-competition, high-intent term should go out before a spoke competing against ten-year-old domain authority sites, because early wins build the internal link equity that helps the harder spokes rank later.
Ready to Turn Your Seed Keyword Into a Published Cluster?
Pick one seed keyword you already rank for on page two of Google, run it through WriteGap's topic cluster planner to generate the hub-and-spoke map with intent grouping and difficulty scores attached, and you'll have a publishing plan instead of a guess by the time you'd have otherwise finished your fourth browser tab.