Measured, not mystical

How NUDGE measures AI visibility and AI opportunity

Trust is earned by being explicit about what is measured, what is inferred, and what cannot be promised. This page explains the AI Visibility Readiness Score, the Observed AI Recommendation Sample, the AI Opportunity Calculator, and the official platform guidance behind them.

AI Visibility Readiness Score

The readiness score evaluates website, local, entity, content and evidence signals under five categories, each worth 20 points of a 100-point total. Every check resolves to one of three states, and every check is explained in the report.

Pass

The signal was verified as present. The check earns its points.

Gap

The signal was verified as missing or weak. The check earns nothing.

Not verified

Not enough could be read to judge it. The check earns nothing either.

That third state matters more than it sounds. An unverified signal never earns credit, because a score that quietly rewards what could not be checked is a score that flatters you. It is measured separately instead, as Assessment Coverage.

Technical Discoverability

20 points

HTTPS available
Homepage is publicly fetchable
Core page is not accidentally noindexed
Robots directives allow core search access
Canonical URL present and plausible
Page title and description are present
Sitemap is discoverable
Important content is available as text
Mobile viewport configured

Entity + Local Clarity

20 points

Business name is clear
Primary city/service area is visible
Contact information is available
About/company page exists
Organization/local business structured data is present when appropriate
Relevant external profiles are linked
Real people/team information is available when relevant
Google Business Profile/local profile can be associated where applicable

Service + Location Relevance

20 points

Primary service is clearly explained
Dedicated service content exists
Location/service-area relevance is explained naturally
Titles/headings accurately describe service
Buyer questions are answered
Relevant internal links connect supporting pages
No obvious city/keyword stuffing pattern

Evidence + Authority Readiness

20 points

Real proof elements are present
Case studies/examples are available when appropriate
Original resources/research/insight exist
Authorship/expertise is transparent
Important informational claims use appropriate sources
External business/profile links are visible
Real reviews/testimonials are represented accurately
Time-sensitive content shows maintenance/freshness where relevant

Answer-Ready Content

20 points

Key services/concepts are defined directly
Process/how-it-works is clear
Who, where, what and why are clear
Important buyer questions are addressed
Differentiation is concrete
Examples/use cases are included
Content includes non-commodity insight
Content is scannable and logically structured

Score bands

Range	Band
85-100	Strong foundation
70-84	Competitive foundation
50-69	Developing
0-49	Significant gaps

Assessment Coverage

Assessment Coverage is how much of the check could be verified from your public website. Signals that could not be verified earn no credit toward the readiness score, so an unreadable site never scores better than a verified one. It is reported alongside the score, never folded into it, so you can always see how much of the assessment actually stands behind the number.

When coverage falls below 80%, there is no overall score at all. The report switches to a partial assessment and says so plainly. You still receive your verified strengths, your verified gaps, the full list of what could not be verified and the recommended next actions. What you do not receive is a confident-looking number built on evidence no one saw. Sites behind aggressive bot protection, sites that render entirely in the browser, and sites that were slow or unreachable during the scan are the usual reasons coverage lands low.

This is an internal diagnostic methodology, not a proprietary score from a search or AI platform.

What the score means

how ready your own site and profiles are to be found, understood and verified by search and AI systems
which of the five categories is strongest and which has the biggest gap
a prioritized list of actions you control
a baseline you can re-run to see whether changes moved the signals

What the score does not mean

it is not a ranking-factor score from Google, OpenAI, Gemini, Perplexity or any other platform
it does not prove your brand is, or is not, recommended in any AI engine for any prompt
a higher score does not guarantee inclusion, ranking or recommendation
it only inspects public business information; nothing behind a login is reviewed

Observed AI Recommendation Sample

The readiness score and the observed sample are deliberately separate. The score looks at signals on your own site. The observed sample, when a sampling provider is enabled, records what AI-assisted platforms actually said in response to a fixed set of commercial queries.

What is recorded

fixed commercial queries, with the exact text saved
provider/platform and model where known
timestamp of the query
whether your brand was mentioned
competitors mentioned
sources and citations where available

How to read it

It is a sample, not a universal rank. An API query does not exactly reproduce what every consumer sees in ChatGPT, Gemini, Perplexity or Google AI Mode, because answers are dynamic and personalized. Client work keeps a fixed benchmark prompt set, repeats it on a schedule, records the platform and model, compares trends and keeps exploratory prompts separate.

AI-generated answers vary by platform, model, query wording, time, location and personalization. This diagnostic identifies discoverability signals and may include observed query samples; it does not guarantee inclusion or ranking.

AI Opportunity Calculator

Inputs are your own operating data plus clearly labeled default assumptions you can change. Outputs are time potentially released, capacity value, and potential revenue opportunity where the inputs logically support one, each presented as conditional sensitivity scenarios that state what would have to be true for the figure beside them. No scenario is presented as the expected outcome, and capacity value and revenue opportunity are never added together.

Capacity value is not cash savings. Capacity value is hours released multiplied by a loaded hourly cost. It becomes a cash saving only when there is a real cash-cost mechanism, overtime avoided, a contractor not hired, a role not backfilled, and the model does not call it “savings” otherwise. Revenue opportunity is computed only from inputs and assumptions you can see and edit; it is not a forecast.

Low scenario

Conservative automation and recovery rates. Use this for planning.

Base scenario

The default assumptions, shown on every result and editable.

High scenario

What a well-adopted system could reach. Not a target and not a promise.

Model notes (version 1.0)

Capacity value is not automatically cash savings.
Revenue opportunity is not a forecast or guarantee.
Use user-supplied operating inputs wherever possible.
Do not infer lost revenue solely from response time.

Estimates are directional and based on information and assumptions you provide. Capacity value is not the same as cash savings, and revenue opportunity is not a forecast or guarantee.

Do not enter customer names, patient information, financial account information, passwords or other sensitive data. Use aggregate business numbers only.

The official sources behind this

Official platform guidance and NUDGE methodology are kept distinct. Nothing here implies that an optimization tactic is officially endorsed unless the source says so, and no access to any proprietary ranking system is claimed.