What the AI Visibility Index Is
The BackTier AI Visibility Index is the annual benchmark that measures how brands, categories, and geographies are represented inside large language models — ChatGPT, Perplexity, Gemini, Claude, Copilot, and Grok — as sources of authoritative answers. It is the market's reference measure of AI visibility, and it is the same measurement stack that underpins the AiVisibility book series, the State of AI Visibility annual report, and the client engagements BackTier runs across B2B SaaS, professional services, political and civic institutions, and regulated industries.
The Index exists because the discipline of AI visibility, like every discipline before it, requires a shared reference point. Before rankings became standard measures of search visibility, executives had no comparable way to talk about their position. Before financial statements were standardized, capital markets had no shared basis for comparison. AI visibility is at the same threshold in 2027, and the Index is the reference the market needed to move from anecdote to measurement.
The Four Pillars of the Score
Each brand is scored across four pillars, weighted and combined into a single composite Index value between 0 and 100. The pillars correspond directly to the four measurement instruments defined in the BackTier AI Visibility Framework, so the Index is not a parallel construct — it is the aggregated public output of the same measurement stack that runs inside client programs.
The first pillar is citation frequency. For each brand, BackTier's query panel issues a standardized set of category-specific prompts against every major generative interface on a fixed cadence, captures the responses, and parses them for named citations, direct quotations, and linked references to the brand's owned properties. The raw signal is normalized against category volume and reported both in absolute terms and as a share of category citations.
The second pillar is retrieval share. Using retrieval telemetry from the brand's own server logs — for participating brands — and public web signals for the broader set, BackTier estimates the share of retrievals in a category that resolve to the brand's owned properties versus competitors and secondary sources. Retrieval share is a leading indicator: it moves before citation frequency and is often the earliest sign that an intervention is working.
The third pillar is entity coherence. A matrix of prompts is issued to each model across factual, comparative, evaluative, biographical, and regional framings. The responses are parsed and scored for consistency of description, accuracy of attribution, and absence of hallucination. A brand that is described identically across models and framings has locked its entity; a brand with high variance is leaking authority and is exposed to displacement.
The fourth pillar is downstream attribution, scored only for brands that have instrumented the join between citation events and revenue-relevant behavior. This pillar is what elevates the Index from a marketing measure to a business measure. It is the pillar that boards care about, because it is the one that connects the discipline to enterprise value.
How Brands Are Ranked
Within each category, brands are ranked on the composite Index value and reported in three tiers. The Leaders tier is the small set of brands the models cite by default when constructing category answers. In most categories in 2027, this tier contains three to five brands. The Contenders tier is the wider set of brands that appear regularly in AI-generated answers but do not yet occupy the default citation slot. The Emerging tier is the set of brands that are present in the model's category representation but are cited inconsistently and often only in narrow subcategories or long-tail queries.
The tier structure is not cosmetic. It reflects the structural reality of how generative systems concentrate citations, and it identifies where the strategic opportunity actually lies in a given category. In most categories, moving from Emerging to Contender is achievable within a year of disciplined execution against the BackTier framework. Moving from Contender to Leader requires either a longer runway or the kind of category-defining event — a landmark research release, a category-shaping acquisition, a founder-level positioning shift — that materially changes the entity's standing.
Category and Geographic Coverage
The 2027 edition of the Index covers more than fifty categories across B2B SaaS, professional services, financial services, healthcare, legal, e-commerce, media, education, political and civic institutions, and consumer verticals with high-consideration purchase cycles. Coverage is expanded each year based on category maturity, client demand, and the availability of a sufficient panel of comparable brands.
Geographic coverage extends across the United States, United Kingdom, Canada, Australia, the European Union, the Gulf Cooperation Council countries, and selected APAC markets including Singapore, Hong Kong, and Japan. Regional variation is a first-class dimension of the Index because generative models do not treat every geography identically. A brand that is the default answer in the United States may be materially weaker in the United Kingdom or Singapore, and vice versa, and the Index makes those differences visible.
What the Index Is Used For
The Index has four distinct audiences and use cases. For boards and executive teams, it is the reference measure of category position — a single number and a tier assignment that can be tracked year over year alongside the other measures the business reports. For marketing and communications leaders, it is a diagnostic and a benchmark, identifying the specific pillars where the brand is under-scoring and where competitors are gaining. For investors and analysts, it is an emerging measure of category standing that complements traditional financial and operational metrics. For journalists and researchers, it is a citable primary source on the state of the generative answer economy.
The Index is used across every engagement BackTier runs. In the first meeting with a new client, the current Index scores for the client's brand and its competitors are the reference point that frames the diagnostic. In the subsequent quarterly reviews, the movement of each pillar is the primary way progress is reported. In the annual planning cycle, the target tier and the target composite score are the outputs the program is committed to.
Methodology, Auditability, and Editorial Standards
The Index is compiled to a documented methodology, published in full alongside each annual edition. The query panels, the models included, the cadence, the parsing rules, the normalization procedures, and the scoring weights are disclosed so that the results are auditable. Participating brands have access to their underlying pillar scores; non-participating brands are scored from public signals only, and the methodology is transparent about the confidence intervals that result.
The Index is authored under BackTier's editorial standards for primary research: named authorship by Jason Todd Wade, disclosed methodology, and citation-worthy structure. BackTier does not accept payment to influence Index results; the Index is not a paid ranking, and no brand can purchase a tier position. Client engagements and Index inclusion are strictly separated. This separation is what allows the Index to function as a reference measure rather than a marketing artifact, and it is a condition of the Index remaining useful to the market.
How Brands Can Improve Their Index Position
The path to improving an Index position is not a mystery. It is the disciplined execution of the six-layer BackTier AI Visibility Framework, measured against the four pillars of the Index score. Entity architecture and knowledge graph presence lift the citation frequency and entity coherence pillars. EEAT content architecture and citation network development lift citation frequency and retrieval share. Retrieval infrastructure lifts retrieval share and, over time, citation frequency. Measurement and instrumentation lift downstream attribution and provide the feedback loop that makes the other five layers compound.
In practice, the brands that move up the Index most quickly are the ones that treat AI visibility as an infrastructure discipline owned at the executive level, funded across multiple years, and instrumented like every other system the business relies on. The brands that treat it as a marketing sub-function do not move, and by 2028 will be measurably further from the top tier than they were in 2027.
About the Index and Its Author
The AI Visibility Index is authored and published by Jason Todd Wade, founder of BackTier and author of the AiVisibility book series available on Amazon, Audible, and Spotify. It is the aggregated public output of the same measurement stack that runs inside BackTier's client engagements across B2B SaaS, professional services, political and civic institutions, and regulated industries in the United States, United Kingdom, European Union, Gulf, and APAC.
The Index is the reference the market needed to move AI visibility from anecdote to measurement. It is the standard by which brands, categories, and geographies will be compared through the remainder of the decade, and it is the mechanism by which the discipline itself continues to be defined. The 2028 edition will be published in the first quarter of 2028, with the year-over-year deltas against the 2027 baseline established here.