PricingSearch articles
Book a strategy call
Link Building

Topical Maps for SEO: Deciding What Not to Publish

A topical map earns its keep by ruling pages out. How to build one, where the boundary sits, and why most maps quietly become a list of things to write.

A subject fanning out to question chips, each carrying a verdict pill, with one marked as a page
Jordan Ellis September 15, 2026 6 min read 1,101 words

A topical map is most useful as a stop list. It tells you which pages a subject genuinely needs, so it’s also telling you which ones it doesn’t.

Topical maps for SEO are usually sold the other way round, as a spreadsheet of forty titles with an implied instruction to write them all.

The version worth building answers a narrower question: what does this subject require, and what is already answered somewhere you own?

What a Topical Map Is

A topical map is a picture of one subject broken into the questions a reader asks, with a decision attached to each one. The decision is the part that gets skipped.

Without it you’ve got a keyword list in a nicer layout. Every node on a useful map carries one of four verdicts: it needs a page, a heading, a section, or nothing.

Four verdict cards showing that only one of them results in a new page

Three of those four verdicts do not produce a new URL. That ratio is the whole point, and it’s why an honest map shrinks a content plan rather than growing it.

How to Build One Without It Turning Into a Backlog

You start from the subject rather than the keyword tool, because a tool hands you every phrase and none of the structure. That’s the wrong end.

Write the questions before the keywords

List what somebody deciding about this subject needs settled, in plain language, in the order they’d hit it. That order is the map’s spine, and it rarely matches search volume.

Attach an existing URL to every question you can

Go through the questions and name the page that already answers each one. Most subjects arrive with more coverage than anyone remembers, scattered across pages written for other reasons.

Give each remaining question a verdict

Only now decide what the gaps need. A question with its own demand and its own answer shape needs a page.

A question that’s a wrinkle on an existing answer needs a heading on that page, and nothing more.

Question chips matched to existing pages, leaving a small group of real gaps

Where the Boundary Sits

You’ll find the hard calls at the edges, and four tests settle most of them.

Test Needs its own page Belongs on an existing page
Intent Someone searching this wants a different outcome Same outcome, narrower wording
Answer shape The answer is a procedure or a comparison of its own The answer is a paragraph
Overlap Under a third of the answer already exists elsewhere Most of it is already written
Linking You can say what would link to it and why Nothing would link to it but the map

The last row catches more bad pages than the other three combined. A page nothing would ever link to is a page nobody asked for.

Same subject, different intent

Two questions can share every keyword and still need separate pages, because the reader wants a different thing from each. Merging on subject alone is how sites end up with pages that compete against each other, and that isn’t a ranking problem you can nudge.

The reverse trap is worse. Splitting one intent across four near-identical pages divides the signals that would have made one page rank.

Why Maps Drift Into Backlogs

It happens because a map with forty nodes looks like more work delivered than a map with six. The artefact rewards volume even when the strategy does not.

Two habits cause most of it. Nodes get created from keyword variants rather than distinct questions, and no node is ever assigned the verdict of nothing.

A map where every node became a page was not a map. It was a content calendar with a diagram attached.

A wide shallow map of keyword variants beside a narrow deep map built from real questions

What the Map Is For After It Exists

The finished map earns its keep as a filter for incoming requests. When somebody proposes a post, the map says whether the subject has a home already.

That’s a faster conversation than opinion, because you’re pointing at a decision rather than arguing about one.

It also tells you where internal links belong, because the structure you drew is the structure a reader moves through. That connection is set out in entity SEO, which covers how clustering reinforces the relationships a search engine reads.

And it shows which pages carry the weight, which is where earned links do the most good. Pointing authority at a hub beats spreading it across leaves, and contextual link building is one route to putting it there.

The Questions a Map Cannot Settle

A map decides what deserves to exist. It doesn’t promise any of it will rank.

Demand and competition sit outside the document entirely, so a node marked as needing a page can still land against three better resources. The map got the structure right and the market decided the rest.

It also says nothing about whether the page is any good. A correct verdict badly executed is still a weak page, and no amount of structure rescues it.

Treat it as a record, not a forecast

What a map gives you is a written reason for each decision, which is worth more six months later than the decision itself. When a page underperforms you can see what it was for.

That’s also how you avoid rebuilding the same argument every quarter, because the reasoning is on the page rather than in somebody’s memory.

Frequently Asked Questions

How large should a topical map be

You cannot know before you build it, and a map that names a target number in advance has decided the answer before doing the work. Most subjects need fewer pages than the first draft suggests.

Is a topical map the same as a content cluster

They overlap. A cluster is a published structure of pages that link to a hub, and a map is the decision document that determines whether those pages should exist. The map comes first.

Do I need a keyword tool to build one

Not to start. The questions come from knowing the subject and the buyer, and a tool is useful afterwards for checking which questions have demand attached.

What do I do with a question nothing on my site answers

Give it a verdict rather than a slot in a calendar. It needs a page only when the intent and the answer shape are genuinely its own, and a heading otherwise.

How often should a topical map be revisited

When the subject moves or the site does. Rebuilding it on a schedule turns a decision tool back into a backlog.

Build the Map to Say No

A map that only ever adds is measuring your ambition rather than the subject. The one worth keeping is the one you can point at when somebody asks why a page does not exist.

Jordan Ellis
Written by

Jordan Ellis

Jordan Ellis is an AI search visibility specialist and content strategist with over 8 years of experience in B2B digital marketing. Focused on the intersection of content strategy and large language model optimization, Jordan writes about how brands can build lasting presence in AI-generated recommendations. Before specializing in AI visibility, Jordan led SEO and content programs for SaaS and FinTech companies across the US and Europe.

Leave a Reply

See where AI answers put your brand today.

Twenty minutes with a senior strategist: where the major engines point buyers in your category, who gets named instead of you, and a straight read on what a programme would change. No pitch deck.

Book a strategy call

A senior strategist replies within one business day.