A keyword export is a pile of strings, not a plan. The obvious approach - sort by volume, take the top 200 - buries you in head terms you cannot win and hides the long tail that actually converts. A topical map fixes the order of publication by grouping demand into entities and sub-entities, so each piece you publish makes the next one easier to rank rather than competing with it.
topical-map-from-demand.md
---
name: topical-map-from-demand
description: Use when you have a subject area but no defensible list of what to publish, and keyword exports are giving you thousands of rows with no shape.
---
# Build a Topical Map From Real Demand
A keyword export is a pile of strings, not a plan. The obvious approach - sort by volume, take the top 200 - buries you in head terms you cannot win and hides the long tail that actually converts. A topical map fixes the order of publication by grouping demand into entities and sub-entities, so each piece you publish makes the next one easier to rank rather than competing with it.
## What you need first
- a keyword export with volume and current position
- your Search Console query export for the last 6 months
- a clear statement of what the site actually sells or does
## Method
1. Dump every query into one sheet, then cluster by SERP overlap rather than string similarity: two queries belong together if the top 10 results share 4 or more URLs. String clustering merges things Google treats separately and splits things it treats as one page.
2. Name each cluster after the entity it is about, not the keyword. "Van insurance for tradespeople" is an entity; "cheap van insurance quotes uk" is a phrasing of it.
3. Tag each cluster with the dominant intent of the current top 5 - transactional, comparison, informational, or navigational. If the top 5 disagree with each other, the SERP is unsettled and the cluster is a cheaper entry point.
4. Mark clusters as core (directly adjacent to what you sell), supporting (answers a question a buyer has on the way), or outer (traffic with no path to revenue). Cap outer at 20 percent of the plan.
5. Order publication so every supporting page has its core page already live, because the supporting page needs somewhere authoritative to link to and it is that internal link that carries the relevance signal.
6. Freeze the map for one quarter and record the date. Re-clustering mid-quarter destroys your ability to tell whether the plan worked or the execution did.
## What this produces
A ranked publication map of named entity clusters, each tagged with intent, tier, and its position in the build order.
## Where this goes wrong
- clustering by keyword string rather than SERP overlap, which produces clusters Google does not recognise
- building the outer ring first because those terms look easiest, so you end up with traffic that never touches a commercial page
- treating the map as a permanent artefact and rewriting it every time a new keyword idea arrives
---
From the QuQi skill library - https://www.quqi.io/skills/topical-map-from-demand
Free to download · no account, no email
A ranked publication map of named entity clusters, each tagged with intent, tier, and its position in the build order.