Topic universe discovery
Find the topic universe before assigning pages.
A topical map is a plan connecting subjects and reader tasks to content homes, relationships and publishing decisions. Discovery supplies candidates for that plan; it does not give every phrase a separate URL.
Free learning from Semantec SEO.
Keep discovery separate from page approval.
A raw candidate list records concepts, questions, entities and tasks that may belong to the site. A processed map decides what to publish, what belongs inside another page and how the pages connect. A calendar then assigns work and dates. These records can support one another without answering the same question.
Retain source context beside the list: the site’s purpose, intended readers, available skills and exclusions. A large universe outside that boundary can create more work without creating a useful site. Start with the source context decision if the boundary is still unclear.
Collect candidates through different routes.
Read original documentation for definitions, parts, attributes, prerequisites and known limitations. Ask relevant readers what they are trying to accomplish and where they get stuck. Record exact questions without assuming one person represents the whole audience.
Use query research to find wording and possible tasks. Add the date and context of any result observations. Compare these candidates with the existing inventory, including useful passages inside broad pages. Search results and competitor pages can reveal leads, but they are not a complete specification of what your site needs.
Worked example: a small bicycle repair specialist.
This hypothetical specialist focuses on routine commuter bicycle maintenance. Its initial candidates include puncture repair, tyre sizing, brake checks and deciding when to seek a mechanic. A component manual may reveal compatibility conditions that generic query research missed. Reader questions may reveal uncertainty about tools or safe stopping points.
Record “repair a puncture” as a task and tyre levers, inner tubes and wheel removal as related candidates. Do not immediately create a page for every item. Some may become sections in one procedure; others may require a separate compatibility explanation or a clear referral to professional help.
Compare the list with current pages. An existing maintenance overview may already answer a candidate well. Mark retain, improve, combine, new candidate or outside scope, with a reason. This example illustrates planning only and supplies no mechanical instructions.
A populated candidate universe and decision record.
Every row below is invented for teaching. The interviews, customer questions, inventory and manual leads did not occur. They show what a research record should contain, not verified bicycle advice or observed search demand. “Need evidence” means the material must still be obtained and checked.
The fictional site explains routine commuter maintenance and refers work beyond its supported scope to a qualified mechanic. This boundary permits questions about routine care while excluding unsupported specialist repair. No row approves a mechanical instruction, a new service or a page for publication.
| Candidate and stable ID | Task and invented origin | Answer evidence and unknown conditions | Proposed home, decision and reason |
|---|---|---|---|
| CT01 · Parts and terms | Reader asks what tyre, tube, valve and rim mean in a puncture discussion. Invented beginner interview lead. | Need a checked distinction between the named parts and the bicycle configurations covered. No component manual has been obtained in this fictional record. | Combine with the existing commuter-care overview as a short orientation. It prepares the reader for later tasks without assuming that every component term deserves a page. |
| CT02 · Puncture repair | Reader wants to understand the full repair process for the bicycle they have. Invented workshop question. | Need an applicable manufacturer or component procedure, model conditions, tools, checks and stop points. Wheel and tyre configuration are unknown; no repair steps are supplied here. | New page candidate: puncture repair. The task may need a complete procedure beyond the broad overview. Hold drafting until the applicable evidence and scope are confirmed. |
| CT03 · Tyre size and fit | Reader asks which replacement tyre can fit their wheel. Invented compatibility question. | Need authoritative sizing references and the relevant rim, tyre, frame-clearance and product conditions. A name or approximate diameter alone is not enough evidence in this record. | New page candidate: tyre compatibility. It owns a selection task rather than a puncture procedure. Link from puncture research only when replacement selection is needed. |
| CT04 · Wheel removal prerequisite | Reader cannot proceed because the repair appears to require removing a wheel. Invented procedure dependency lead. | Need the bicycle and axle/brake arrangement, applicable instructions and scope of a beginner procedure. This fictional record does not establish that removal is required in every repair. | Defer as a separate page. First check if the supported puncture procedure can explain the prerequisite accurately. A distinct page needs its own complete evidence and task. |
| CT05 · Tools and stop conditions | Reader asks which tools are needed and when to stop and ask a mechanic. Invented support question. | Need tools tied to the supported procedure, conditions outside the instructions and a clear professional referral route. Do not invent a universal safe repair boundary. | Improve the puncture candidate brief: put tool needs and stop conditions beside the relevant steps. Keep a plain referral path in the broad overview too; no standalone tool list is justified yet. |
| CT06 · Brake problem diagnosis | Reader reports a braking problem after routine commuter use. Invented customer question. | Need the brake type, observed symptoms, applicable documentation and qualified review. No diagnosis, repair instruction or permission to ride is established by the question. | Defer the troubleshooting page and provide a qualified mechanic contact route in the plan. The specialist must confirm scope and evidence before this becomes a procedure. |
| CT07 · Existing commuter-care overview | The fictional inventory contains one broad orientation page covering routine-care decisions. Invented inventory observation. | Need to read the complete existing page, check its facts and find gaps against the accepted reader tasks. Its title alone does not establish that any candidate is already answered. | Retain this overview as the navigation home, then improve its route to detailed answers. Do not replace it with a second broad page using a different title. |
| CT08 · Carbon racing-frame repair | A query lead proposes repairing a competition carbon frame. Invented query candidate; no search volume or result study exists. | Need specialist evidence and competence outside this fictional commuter-maintenance remit. No such service or review owner has been established. | Exclude from this site plan with the reason: outside the agreed subject and available evidence. A commercial-looking query is not permission to expand the site into an unsupported repair field. |
| CT09 · Replacement inner-tube wording | Another invented question asks how to identify the correct replacement tube. It resembles CT03 but focuses on a particular component. | Need applicable tube, tyre and valve information, with the conditions and terminology verified. The record has no evidence that a distinct page is useful. | Combine provisionally into CT03’s compatibility brief. Check that the passage can answer the tube task fully; split later only if research establishes a distinct answer and enough evidence. |
CT02 and CT03 remain separate candidates because carrying out a supported procedure and choosing a compatible replacement can require different answers. CT09 is provisionally combined with CT03 because the record has not established enough difference for another page. That combination must still leave the tube question fully answered; a convenient grouping is not a reason to omit it.
CT04 and CT06 are deferred for different reasons. Wheel removal is an unresolved prerequisite whose correct home depends on the eventual procedure. Brake diagnosis lacks the qualified evidence and review needed even to define a useful troubleshooting scope. CT08 is excluded because it conflicts with the site boundary, rather than merely waiting for a convenient keyword.
The record exposes a blind spot: reader questions suggest beginner needs, while the proposed compatibility answers require component-specific documentation that has not been acquired. The next research work is to obtain suitable original instructions, verify the relevant configurations and ask a qualified reviewer to define the support boundary. It is not to write confident procedures from these invented leads.
A real project would replace each invented origin with a retrievable source and date, retain the candidate ID through grouping, and record who accepted the decision. Retain, improve, combine, new candidate, defer and exclude are proposed editorial states here. They are not observations of page quality, search results or completed review.
Look for the blind spots in the collection method.
A query export can miss unfamiliar problems that people cannot name. Existing customers can overrepresent experienced users. Competitor pages can repeat the same omissions. Original documentation can explain a product well but leave a beginner’s decision unresolved.
Use a coverage check across definitions, decisions, prerequisites, procedures, exceptions, failure states and support. Record empty categories as questions for research rather than inventing content to fill a symmetrical map. Add a confidence or evidence state to every candidate so uncertainty survives later grouping.
Give candidates stable records.
A usable discovery record contains a stable ID, subject, reader task, source, scope fit, existing content home, unresolved questions and proposed next action. Keep the original observation distinct from your interpretation. Different phrasings may point to one answer; similar wording may conceal different conditions.
Review overlaps before assigning a URL. A page earns its own home through a distinct useful task and enough evidence to answer it, not through the presence of a phrase in a spreadsheet. Preserve links to the source record when candidates are combined.
Review the candidates before page decisions.
Use the worksheet below to record the evidence in your own project. Discovery is followed by grouping, overlap review and page ownership decisions. Keep proposals separate from approved pages.
The topical mapping hub connects the process. Return to the learning overview for another subject. Google’s helpful content guidance offers an external usefulness check; it does not prescribe a topical map format.
Candidate topic worksheet
Copy these fields into your own notes or spreadsheet. This is a static worksheet.
| Field | Record |
|---|---|
| Candidate ID | Stable reference retained during grouping |
| Subject and task | What the reader needs to understand or do |
| Origin | Original source, reader question, query or inventory observation |
| Evidence state | Observed information and unresolved assumptions |
| Current home | Existing page or passage that already answers the need |
| Next action | Research, combine, improve, consider a new page or exclude |