Search Intent to Content Type and Format Map
Turn classified intent into the specific page format, structure and assets each keyword needs.
Takes keywords whose intent is already known and maps each one to the content type, page structure, required on-page assets and success signal — the production decision that follows intent classification rather than the classification itself.
Ready-to-use prompt
The prompt
Copy it as-is, then swap the bracketed placeholders for your own details before running it.
Role: You are a content strategist converting search intent into production specs.
Context:
- Keywords with known intent: {{keywords_with_intent}}
- Content formats we can produce: {{available_formats}}
- Team capacity per month: {{capacity}}
Task: For every keyword, return:
1. Content type (guide, listicle, comparison, template, tool page, glossary entry, product page)
2. Required page structure (the H2s a page of this type needs to satisfy the intent)
3. Mandatory on-page assets (table, calculator, screenshots, downloadable, example output)
4. Approximate depth (short, standard, deep) and why
5. The success signal to judge it by (rankings, conversions, assisted conversions, internal navigation)
Rules:
- Only propose formats from the available list. If the intent needs a format we cannot produce, say so and give the closest viable substitute.
- Do not re-classify intent — take the supplied intent as given.
- Never assign the same structure to two different intents.
Output: a production map table, then the keywords whose intent our formats cannot serve.
- Mark any asset dependency that could exceed stated capacity; do not assume specialist support.
- Define one observable event or metric for each success signal.
- Separate formats that fully satisfy intent from substitutes that leave the core action unsupported.
Estimated results
Editor's note
Why this prompt matters
Knowing a keyword is commercial does not tell a writer what to build. The gap between intent and production is a format decision, and teams that skip it publish the same 1,200-word article shape against every intent. This prompt takes intent as an input and returns the page type, the headings that intent requires, the assets that must appear on the page, and the metric that will judge it.
Anatomy
Prompt engineering breakdown
Role
Context
Goal
Constraints
Output format
What you'll get
Expected output
Worked example: a capacity-planning resource library
A small editorial team can produce guides, comparisons, templates and glossary entries, with capacity for two standard pages per month. Four keywords have already been selected. The SERP observations below are illustrative inputs, not claims about live results. Classification happens upstream; the production map accepts those labels unchanged.
Classifying intent before you map
Before running the mapping prompt, inspect current results for each keyword in the target market. Record dominant page types, titles, snippets and the action landing pages enable. Open representative results rather than treating a word such as “best” or “download” as conclusive evidence.
- “how to calculate team capacity” — informational: suppose results predominantly explain available hours, allocations and formulas. The searcher needs a method and worked answer.
- “capacity planning template vs software” — commercial: suppose results compare setup effort, maintenance and planning capabilities. The searcher is evaluating approaches before committing.
- “capacity planning template download” — transactional: suppose results lead directly to usable spreadsheet downloads. Obtaining the asset, rather than reading about templates, is the immediate goal.
- “{{product_name}} login” — navigational: suppose results point to the named product’s sign-in destination and access support. The searcher wants a particular location.
If results are mixed, document the competing evidence and have a strategist approve the primary label before mapping. Preserve that decision alongside the keyword; do not ask this production prompt to silently resolve it.
Production map
| Keyword | Type and required H2s | Mandatory assets | Depth and reason | Success signal | |---|---|---|---|---| | how to calculate team capacity | Guide: Calculate available hours; Subtract commitments; Work through an example | Formula, assumptions table, completed calculation | Standard: enough detail to reproduce the calculation | Rankings for the supplied query | | capacity planning template vs software | Comparison: Compare planning needs; Assess maintenance; Choose an approach | Decision matrix, equivalent worked scenarios | Standard: support a defensible choice | Assisted conversions to the template download | | capacity planning template download | Template: Download the spreadsheet; Enter team inputs; Check your output | Downloadable file, preview, completed example | Short: prioritise access and essential instructions | Conversions: completed downloads | | {{product_name}} login | Guide, closest available but inadequate: Find sign-in; Recover access | Verified destination link, access instructions | Short: troubleshooting only, not a replacement destination | Internal navigation to verified sign-in |
Capacity decision and unmet intent
Produce the calculation guide and template first as a connected pair; defer the comparison. This sequence assumes each fits one standard production slot, including asset creation and review.
Unsupported keyword: “{{product_name}} login.” A guide cannot replace the expected sign-in endpoint. Do not commission it merely to fill the map. Verify whether an existing destination can serve the query and route the request to its owner.
Under the hood
Why this prompt works
Constraining the model to your real format list is what makes the output usable. It converts an open-ended creative answer into an allocation problem against known capabilities, and the rule that no two intents share a structure prevents the quiet collapse back into one generic article template.
Model fit
Best AI models for this prompt
Claude
Best at holding the constraint that two intents must not receive the same structure.
ChatGPT
Fastest for large maps and good at proposing asset ideas within the format list.
Gemini
Strong when you paste two competing pages per intent so it can infer the structure the SERP rewards.
When to use
- After intent labels are approved, before drafting briefs.
- When writers keep giving different search tasks the same article structure.
- When allocating a quarter’s approved keywords against fixed production capacity.
- When checking whether planned assets actually support each page’s intended action.
When not to use
- Before intent is known; inspect SERPs and classify first.
- When deciding which keywords merit separate pages; prioritise and consolidate first.
- For campaign strategy, audience selection or channel allocation.
- When live SERP evidence, analytics or engineering estimates are missing; obtain those inputs rather than accepting guesses.
Get more from it
Pro tips
- 1
Supply approved intent labels and retain their evidence separately; this prompt maps production, not intent.
- 2
List buildable formats and asset constraints, including whether spreadsheets or calculators need developer support.
- 3
Turn required assets into acceptance checks: test downloads, reproduce calculations and verify screenshots before approval.
- 4
Re-run when available formats change; compare substitutions against the previous map.
- 5
Define capacity in page-equivalents or hours so a deep guide does not count like a short glossary entry.
- 6
Flag identical H2 sets across different intents in a second pass; request task-specific replacements.
- 7
Challenge every substitute: does it enable the intended action, or merely discuss it?
Don't ship this
Common mistakes
✗ Letting the prompt re-classify intent it was given.
Fix — State that intent is fixed input and reject any output that changes it.
✗ Accepting formats the team cannot produce.
Fix — Provide a closed format list and require substitutes.
✗ Ignoring the success signal column.
Fix — Set it before publishing, otherwise every page is judged on rankings alone.
People also ask
Frequently asked questions
Q.Does this classify search intent?
No. It takes intent as an input. Use the search intent classification prompt first, then feed its output here.
Q.What if our formats cannot serve an intent?
The prompt returns those keywords separately with the closest viable substitute, which is usually the strongest argument for building a new format.
Q.Can I use it for an existing page?
Yes — compare the returned structure against the live page and treat the differences as the rewrite brief.