The AI Visibility Scorecard: how to measure whether AI can find, trust, and cite your firm

You cannot manage what you do not measure. That old rule is the whole reason the AI Visibility Scorecard exists. When a prospect asks ChatGPT, Perplexity, or Google AI Overviews to recommend a firm in your category and market, one of two things happens: your name comes up, or it does not. Most firms have no idea which, and no way to check. The scorecard from Aday Interactive, Inc. gives you an objective read on whether AI engines can find you, trust you, and cite you by name, so you can fix the right thing instead of guessing.

The short version

The metrics most firms still watch measure a buyer behavior that is fading. What matters now is how often an AI engine names your firm when someone asks for a recommendation, and that comes down to three things you can measure: can the engine find you, does it trust you, and is your content shaped to be cited. Grade those three, read your band, fix the lowest dimension first, and re-check quarterly. That is how a firm moves from Invisible to Citable, and it is exactly what our free GEO Audit at /geo-audit measures for you.

This matters most to the people who own the number: CMOs, managing partners, and growth leads at professional-services firms. You already track something. The question is whether you are tracking the thing that now drives new business, or a proxy for it that stopped meaning much the moment buyers started asking an AI engine for a shortlist instead of scrolling a page of links.

Why the old metrics quietly stopped mattering

Pageviews, impressions, and rankings all measure the same buyer behavior: a person types a query, sees a list of links, and clicks one. That behavior is shrinking. More buyers now read a synthesized answer at the top of the page, or never touch a search engine at all and just ask an assistant directly. In both cases there is no list to rank in and no link to earn a click. There is one answer, and either your firm is named in it or it is not.

This is why a healthy traffic report can sit right next to a quiet pipeline. The traffic is real, but it is measuring the wrong surface. The metric that matters now is your share of AI recommendations: out of all the times an engine answers a buying question in your category, how often does it name your firm. That number does not show up in a standard analytics dashboard, because the engine never sends a visit you can count. The scorecard measures the inputs to that number instead, the signals an engine reads before it decides who to name, so you can move them on purpose rather than hope traffic converts.

The three dimensions the scorecard grades

An AI engine has to clear three separate bars before it will cite a firm. It has to be able to read you. It has to believe you are who you say you are. And it has to find your content in a shape it can lift into an answer. The scorecard grades each of these as its own dimension, because a firm can be strong in one and failing in another, and the failing one caps the rest. We call them Findability, Trustworthiness, and Citation Probability.

Dimension 1: Findability

Findability asks a blunt question: when the crawler behind an AI engine fetches your pages, does it receive real text, or an empty shell. The crawlers that feed these answers, including GPTBot, PerplexityBot, and ClaudeBot, download your HTML and largely do not run JavaScript. A site built as a browser-rendered single-page app can look flawless to a human and arrive nearly blank to the crawler, because the words are painted after load. If the engine cannot read you, nothing else on the scorecard can save you, which is why this dimension comes first.

Findability also grades your JSON-LD schema completeness and structure. Schema is the machine-readable label that tells an engine exactly what your firm is, what it does, and how its people and services connect. We check whether the right types are present for your field, whether they are wired together with identifiers into one connected entity rather than scattered fragments, and whether the sitemap and robots file actually point crawlers at the pages you want read. A high Findability score means the raw file the crawler downloads contains your practice areas, your location, your people, and clean structured data, all before a single line of JavaScript runs.

Dimension 2: Trustworthiness

Being readable is not the same as being believed. Trustworthiness grades whether an engine has reason to treat your firm as a real, verified entity worth staking an answer on. The first input is entity verification: does your firm name, category, address, and the credentials of your people line up across your own site, your licensing or bar directory, your Google Business Profile, and the major review platforms. Every mismatch, a middle initial here, an old suite number there, a different practice-area label somewhere else, is a small reason for the engine to doubt that all these listings describe one firm.

The second input is citation consensus, which is agreement across independent sources. When several sources the engine did not have to take your word for say the same thing about your firm, its confidence climbs. When they conflict or go silent, it hedges by naming someone else. The third input is review sentiment: not just a star average, but a reading of what verified reviewers on independent platforms actually say, handled within the rules that apply to your field, whether that is bar advertising guidance, healthcare review constraints, or the SEC framework governing testimonials for advisers. Trustworthiness is the dimension that most often separates two firms with equally clean websites.

Dimension 3: Citation Probability

The last dimension asks whether your content is shaped so an engine can actually quote it. Citation Probability grades answer-formatting first: does an important page lead with the answer to a real question in its opening sentence, or bury it three paragraphs into a story. Engines lift the clean, direct passage. A page that opens with a line about your commitment to excellence gives the engine nothing to extract. A page that opens by answering the exact question a prospect typed hands it a sentence it can cite.

It also grades content structure and E-E-A-T signals, the evidence of Experience, Expertise, Authoritativeness, and Trust that engines lean on for high-stakes queries. That means question-phrased headings in the words a buyer types, specific and verifiable claims rather than vague ones, named authors with real credentials, and bios connected to licensing records through schema. When most of the content on a topic is unsigned and generic, a bylined piece from a named, credentialed practitioner stands out as the source an engine can trust. Citation Probability rewards content written to be extracted and attributed, not just admired.

How to read your score: Invisible, Emerging, Citable

The three dimensions combine into one number on a 0 to 100 scale, sorted into three bands. A score of 0 to 40 is Invisible. The engine cannot reliably read, verify, or cite you, so a buyer asking for a shortlist in your category almost never hears your name, and every dollar of marketing spend leaks out the bottom of a broken funnel. Firms in this band usually have one dimension near zero, most often Findability, and fixing that one thing can move the whole score fast.

A score of 41 to 75 is Emerging. You appear in AI answers sometimes, inconsistently, depending on how the question is phrased. Lead flow is real but unpredictable, and small, targeted fixes tend to move you quickly because the foundation is mostly there. A score of 76 to 100 is Citable. Engines can find you, trust you, and recommend you by name, and appearing on a shortlist becomes the normal case rather than a lucky one. The work then shifts from building to defending, because competitors keep improving and the engines keep raising the bar. The value of the bands is focus: you fix the lowest dimension first, because it is the ceiling on the rest, then re-measure to see what moved.

The score reads differently by vertical

The same 0 to 100 scale applies everywhere, but the effort behind a good score is not evenly distributed across fields, because the trust bar is not the same everywhere. For a law firm, Findability and clean bar-compliant content tend to be the swing factors, and consistency across your bar directory profile carries real weight. Getting the schema and answer-formatting right often moves a legal firm out of the Emerging band without heroic effort.

Medical and wealth firms face a higher floor on Trustworthiness. When a buyer asks an engine to recommend a concierge physician or a firm to manage a family's money, the stakes of a bad answer are high, so the engine leans harder on credential verification and citation consensus before it names anyone. That means entity verification and independent, rule-aware review signals do more of the work, and a thin or unverifiable trust profile will cap a medical or wealth score even when the website itself is excellent. Knowing where your vertical's weight sits tells you which dimension to invest in first, which is the whole reason to measure before you spend.