What it is
A fast diagnostic that shows whether your content is built to satisfy the range of intent types modern search and answer systems classify and route. It helps you spot mixed-intent pages, missing intent coverage, and the specific fixes that make pages easier to select and use in the answer layer.
Who it’s for
SEO leads, content strategists, and marketing owners who want a practical way to improve AI-era visibility without rewriting everything.
When to use it
Use this when:
- You rank in classic search but rarely show up in AI answers
- You show up inconsistently across answer engines or runs
- Your content is “good” but gets ignored, not cited, or not used
- Your site has grown over time and templates have drifted into sameness
What you’ll need
- A list of your top pages. Start with 10. Expand to 25 or 50 once you’ve run it once.
- A place to take notes. A word doc works fine.
- A timer. This is meant to be fast.
The intent menu
Most teams plan for three. Modern systems operate with many more. You do not need to use every label below. You do need to design pages with a clear primary job.
Core intents
- Informational. The user wants to understand.
- Navigational. The user wants a specific site, brand, or page.
- Transactional. The user wants to buy, book, subscribe, or contact.
Task intents
- Procedural. The user wants steps to complete a task.
- Comparative. The user wants options, tradeoffs, alternatives, best vs. X
- Diagnostic. The user wants to identify cause, problem, or fix.
- Evaluative. The user wants to validate a short list or confirm a decision.
Context intents
- Local. The user needs nearby, service area, availability, hours.
- Freshness. The user wants latest, updated, current, today.
- High-stakes. The user needs safe guidance where bad answers carry risk.
- Compliance. The user needs official requirements, eligibility, rules, definitions.
Experience intents
- Exploration. The user wants to browse examples, ideas, inspiration, collections.
- Reassurance. The user wants confidence, credibility, a second opinion, proof.
- Troubleshooting. The user wants a fast fix, not a deep explanation.
Mixed intent
- Some queries legitimately contain more than one intent. Your page still needs one primary intent. Secondary intents can be supported, but they cannot compete with the primary job.
The 20-minute audit
Run this on 10 pages first. It will expose patterns quickly.
Step 1: Pick the 10 pages
Choose pages that matter. Use a mix like:
- 3 revenue pages
- 3 high-traffic pages
- 2 decision-support pages, like comparisons or pricing explainers
- 2 “authority” pages, like a flagship guide or core resource
Step 2: Assign one primary intent per page
For each page, write one line:
- This page exists to help someone do X.
Then label the primary intent type from the menu above.
If you cannot do this in one line, mark the page as mixed-intent.
Step 3: Check first-screen intent clarity
Scroll to the top of the page. Do not skim. Ask one question.
- Does the page make its job obvious in the first screen?
Score it:
- Clear. The job is obvious without scrolling.
- Delayed. The job becomes obvious after scrolling.
- Unclear. The page never commits, or multiple jobs compete.
Step 4: Check for the quote-safe block
Find the one block of content you’d want an answer engine to quote.
A quote-safe block includes four elements:
- Direct answer. A clean statement that stands alone.
- Boundary. When it applies and when it doesn’t.
- Proof. A source, data point, example, or constraint that grounds it.
- Next step. The action a user should take immediately after.
If you cannot find a quote-safe block, mark the page as hard to use.
Step 5: Run the mismatch test
Look for these common mismatch patterns:
- One page trying to teach and sell at the same time
- A how-to page that never becomes a how-to
- A comparison page with no criteria
- A local page that reads generic
- A high-stakes page making absolute claims
- A transactional page hiding the next step
Step 6: Choose the fix, only three options
- Refocus – Keep the page, pick one job, remove or move competing sections.
- Split – Turn one mixed-intent page into two pages, each with a single primary intent.
- Build – Create a net-new page when you have intent demand but no clear page built for it.
How to read the results
This framework is designed to surface decisions quickly.
- If 3 or more pages are mixed-intent, you have an intent clarity problem.
- If 5 or more pages lack a quote-safe block, you have a usability and trust problem in the answer layer.
- If your 10 pages cover only the three core intents, your content is built for a narrower discovery model than modern systems use.
How to expand it beyond 10 pages
Once you’ve run the 10-page pass, scale it to 25 or 50 pages and group them by funnel stage. You’ll usually find the same gap patterns repeat:
- Too many pages competing on informational intent
- Too few pages built for comparison and validation
- Local pages that don’t prove locality
- High-stakes content that lacks boundaries and evidence trails
What success looks like
After fixes, you should be able to say:
- Every high-value page has one primary intent and it’s obvious fast.
- Every high-value page has at least one quote-safe block.
- Pages built for comparison have criteria.
- Pages built for local intent prove local service boundaries.
- High-stakes pages use precise language, boundaries, and sources.
The question to ask your team
How many intent types are we intentionally designing for, and where are we still forcing the system to guess?
Feel free to use this framework, expand on it, etc. It’s meant as a starting point for your work.