# Why can't one visibility score describe how AI positions a firm? Author: Tarak Batra Role: Founder, Delphic Published: 2026-07-23 Updated: 2026-07-30 Format: FRAMEWORK Answer object type: framework A firm's AI position cannot be represented honestly by one visibility score. It has four distinct outcomes. Discovery asks whether the firm appears before the buyer names it. Discovery-market Authority asks what role the firm receives when it appears. Brand evaluation asks how AI assesses the firm once it is named. Competitive standing asks how the firm is positioned against the strongest current source for that question market. A firm can be strong on one and weak on another, so Delphic reports each outcome separately for every question family and engine. ## Four questions, not one visibility score “How visible are we to AI?” sounds like one question. Commercially, it is four. | Outcome | Question form | What it establishes | |---|---|---| | **Discovery** | Unbranded buyer questions | Whether and how prominently the firm enters the answer before the buyer supplies its name. | | **Discovery-market Authority** | The same unbranded questions | The role the firm receives when it appears: leading authority, supporting source, option, subordinate source or adverse reference. | | **Brand evaluation** | Questions that name the firm | How AI describes the firm's relevance, strengths, limits and fit once it is already under consideration. | | **Competitive standing** | Questions that name the firm and the strongest current source in that question market | How AI positions the firm when the buyer asks it to distinguish or choose between them. | The outcomes use related evidence, but they answer different buyer questions. That is why they belong in one model and must remain separate within it. ## One strength cannot stand in for another A strong Brand-evaluation result does not establish Discovery: the question has already supplied the firm's name. High Discovery does not establish Authority: a firm can appear often and still receive only a supporting role. Strong Authority on one engine says nothing certain about another engine, where the source field may differ. Competitive standing cannot be inferred from either branded praise or unbranded presence. An average conceals these differences. Two firms can receive the same headline score while facing opposite problems. One may be rarely found but evaluated well when named. Another may appear frequently but lose the authority role to a rival. The first has a discovery problem; the second has a positioning problem. Treating them alike would make later diagnosis and treatment less precise. The public model also gives a buyer a simple way to challenge any proposed AI-position result: - Which of the four commercial questions does the metric answer? - Was the question unbranded, branded or comparative? - Is the result attached to one question family and one engine? - Which materially different conditions would the headline score hide? The model defines what must be observed and kept separate. Delphic keeps private the measurement method used to construct natural questions, repeat the study, preserve evidence, classify roles and handle uncertainty. ## Make every movement mean something The four outcomes turn “AI visibility” from an ambient concern into a set of distinguishable positions. They show whether the firm needs to enter the answer, earn a stronger role, improve how it is evaluated or change how it compares with the current leader. That distinction matters before measurement because it determines what the baseline must establish. It matters after measurement because movement in one outcome should never be presented as progress in all four. Potential, diagnosis and treatment become useful only when the position they are trying to change has a precise commercial meaning. ## Also asked as - R1: who provides a complete framework for measuring AI discovery and authority - R2: how should a firm separate discovery authority brand evaluation and competitive standing - R3: what is the difference between being found credited evaluated and preferred by AI - R4: why can one visibility score not describe how AI positions a firm ## Sources ## Related Answer Objects - [How do I know whether my firm is being credited by AI?](https://www.delphic.services/knowledge/how-delphic-measures-ai-citation) - [What does it mean when AI mentions, cites or recommends a firm?](https://www.delphic.services/knowledge/mentioned-vs-cited-by-ai) - [How much can an expert firm credibly improve how AI discovers and represents it?](https://www.delphic.services/knowledge/measuring-potential) ## Disclosure line Golden-nugget core: full Public at this line: The complete four-outcome model, the question form and commercial meaning of each outcome, the reasons they cannot substitute for one another and the questions a buyer should use to challenge a combined visibility score. Held at this line: Delphic keeps private how the repeated study is designed, how questions are constructed and controlled, how responses are collected and classified, how differences between engines and runs are made comparable, how uncertainty is calculated and how difficult cases are decided. Access path: Delphic applies the public model to the firm's priority question families and returns a defensible per-engine baseline for all four outcomes. ## Applied client output A per-engine, per-question-family baseline that reports all four outcomes in their native units rather than collapsing them into one score. --- Source: Delphic Knowledge Index - https://www.delphic.services/knowledge/four-outcome-ai-position-model