CLEAR is a five-dimension standard for scoring whether a web page works for both audiences that now read it: the human buyer and the AI agent. Each page is scored 0 to 100 on Credibility, Leverage, Evidence, Alignment, and Responsiveness, from both perspectives. The gap between the two scores is your work list.
Free for ten pages. No credit card, no HubSpot account needed.
What are the five CLEAR dimensions?
Each dimension carries two labels because it is scored twice: once for the human question it answers, once for the signals an AI agent reads. Below, each dimension lists the human question, the agent signals Audaeo checks, the metrics it extracts, and what a good page looks like.
C
Credibility · Clarity
Human question
Can a person trust this page and grasp the offer in one screen?
AI-agent signals
Valid JSON-LD (Organization, Product, FAQPage), one H1 with logical H2 and H3 nesting, semantic HTML (main, article, nav), HTTPS, and a self-referencing canonical. This is the layer that decides whether an engine can parse the page at all.
What a good page looks like: A product page whose claims are marked up in schema, so an engine lifts them verbatim without guessing.
L
Leverage · Logic
Human question
Does this page tie to a business outcome, not a vanity metric?
AI-agent signals
A clear logical structure an engine can follow: cause and effect stated plainly, comparison tables, defined entities, and internal links that connect related concepts. Numbers stated as standalone facts rather than buried in prose.
Extractable metrics
Named metricsTabular dataEntity linksInternal-link graph
What a good page looks like: A pricing page that states tiers and prices in a table an engine can compare, not a PDF or an image.
E
Evidence · Emotion
Human question
Is every claim backed by proof a reader can check?
AI-agent signals
Structured proof: Review or Rating schema when it is real, cited primary sources with outbound links, datePublished and dateModified, and author markup. Named figures instead of adjectives. Language that reads measured and specific, which engines weigh over hype.
What a good page looks like: A case study with a named result and a linked source, marked up so an engine can attribute it.
A
Alignment · Accessibility
Human question
Does this page fit where the buyer is, and does the site cover the whole journey?
AI-agent signals
Content mapped to a funnel stage, topic-cluster internal linking from pillar to cluster and back, BreadcrumbList schema, a sitemap.xml, an llms.txt, a robots.txt that allows AI crawlers, and HTML that is crawlable without JavaScript.
What a good page looks like: A pillar page that links down to every supporting article and back up, with a sitemap and llms.txt an engine can follow.
R
Responsiveness · Relevance
Human question
Does this page answer the exact question the buyer asked?
AI-agent signals
Answer-first structure with the answer in the first 40 to 60 words, question-based H2 and H3 that mirror real prompts, FAQPage schema, concise extractable blocks, and coverage of the sub-queries the query fan-out generates.
What a good page looks like: A page that opens with a two-sentence answer to the question in its H1, then supports it.
How does CLEAR map to Why, How, and What?
CLEAR covers all three altitudes of Simon Sinek's Why, How, and What, which line up with the funnel. Most audits only speak to What. CLEAR scores the purpose and the method too, so a page earns trust before it makes the pitch.
Funnel stage
Sinek
CLEAR emphasis
What the page says
Awareness
Why
Credibility, Alignment
The buyer journey now runs through one AI answer. Most sites are not in it.
Consideration
How
Leverage, Responsiveness, Evidence
Every page, scored from a human's view and an agent's view. The gap is your work list.
Decision
What
Credibility, Evidence, Leverage
Specs, price, security, and proof, stated as facts an engine can quote.
How is a CLEAR score calculated?
Every page is scored 0 to 100 across the five dimensions, from both the human and the AI lens, then rolled into one number and a letter grade. Page scores aggregate into a site grade, so one weak page shows up as a named issue rather than hiding in an average.
beaconfleet.com
C
0/100
AEO Readiness
Human AI agent
Credibility · Clarity
H
82
AI
68
Leverage · Logic
H
70
AI
61
Evidence · Emotion
H
66
AI
52
Alignment · Accessibility
H
79
AI
58
Responsiveness · Relevance
H
74
AI
63
beaconfleet.com/product/dispatch
Dispatch & Route Optimization
D
0/100
Credibility
6244
Leverage
5438
Evidence
4834
Alignment
6042
Responsiveness
5640
Top issues
Critical: No FAQ or Product schema — engines can't extract answers
Warning: Zero customer proof or named results on the page
Warning: Core specs locked inside an image, invisible to crawlers
Grade
Score
What it means
A
85–100
Reads clean to a human and parses clean to an agent.
B
70–84
Solid, with a few extractability gaps to close.
C
55–69
Human-legible, but engines struggle to parse or cite it.
D
40–54
Real machine-readability blockers are dragging the page down.
F
0–39
An engine can barely read or trust the page.
Common questions about CLEAR
Where does the CLEAR framework come from?+
CLEAR was developed by Impulse Creative, the agency that builds Audaeo, and is documented in its CLEAR framework ebook. Audaeo turns the framework into a scored audit and holds its own marketing site to the same standard.
What is the difference between the human score and the AI score?+
The human score rates how a person reads the page: trust, clarity, and whether the offer lands. The AI score rates how an answer engine parses it: schema, structure, extractable facts, and crawlability. A page can score well on one and poorly on the other. The gap is what you fix first.
Does every page get its own CLEAR grade?+
Yes. Each page is scored 0 to 100 across the five dimensions from both perspectives, then rolled up into a site grade. This is why one weak page, like a product page with no schema, can pull the whole domain down and show up as a named issue.
What CLEAR grade should I aim for?+
Aim for an A, 85 or above, with no page below a B. That is the launch gate Audaeo holds its own site to. A B means solid with a few extractability gaps; a C or below means engines struggle to parse or cite the page.
How is CLEAR different from a traditional SEO audit?+
An SEO audit scores ranking factors: keywords, backlinks, and page speed. CLEAR scores whether an AI engine can discover, understand, and cite the page, from both a human and an agent lens. Different metric, different outcome: recommendation inside the answer, not a ranked link.
Get your CLEAR grade
Score ten pages across all five dimensions, from both a human and an AI-agent lens. Free.
