BrandGEO The Authority Waterfall: The Upstream Model of AI Visibility — BrandGEO A Markdown version of this page is available at https://brandgeo.co/blog/authority-waterfall-ai-visibility-upstream-credibility.md, optimized for AI and LLM tools. [ AI Visibility ](https://brandgeo.co/blog/category/ai-visibility) [ Brand Strategy ](https://brandgeo.co/blog/category/brand-strategy) · April 19, 2026 · 9 min read · Updated Apr 23, 2026 The Authority Waterfall: Why AI Visibility Flows From Upstream Credibility ============================================================================ LLMs don't build a brand's credibility — they reflect it. Here's the upstream model that explains what actually moves your score. The first time a marketing team runs an AI visibility audit and sees a disappointing score, the reflex is almost always the same: what do we change on our site to fix this? Schema markup, structured data, better on-page content, a clearer about page. All of those are reasonable instincts. Most of them are also wrong — not because they do not matter, but because they operate downstream of the actual cause. This post introduces a framework we call the Authority Waterfall: the model that explains where AI visibility actually comes from, and why the fix is rarely on the page that fails the audit. The first time a marketing team runs an AI visibility audit and sees a disappointing score, the reflex is almost always the same. What do we change on our site? Which meta tags need updating? Should we add schema? Should the about page be clearer? All reasonable instincts. Most of them are also wrong — not because they do not matter, but because they operate downstream of the actual cause. AI visibility is not primarily made on your site. It is primarily made upstream of your site, by the ecosystem of citations, mentions, and credibility signals that language models used to learn about your category in the first place. This post introduces a framework we call the Authority Waterfall. It is the mental model that explains where AI visibility actually comes from — and why the fix is rarely on the page that fails the audit. What the waterfall is --------------------- The Authority Waterfall describes how credibility signals flow from external, third-party sources down through increasingly proprietary surfaces, eventually arriving at the AI answer a buyer sees when they ask about your category. The layers, from top to bottom: **1. Editorial authority.** Coverage in widely-read, credibility-conferring publications. For B2B, this means HBR, McKinsey, industry trade press, major newspapers. For consumer, it extends to lifestyle and vertical publications. Editorial coverage is the single highest-signal input to how language models assess a brand's category position. **2. Analyst and review authority.** Gartner reports, Forrester waves, G2 / Capterra / Trustpilot entries, industry-specific analyst coverage, vertical review aggregators. These are the sources that language models disproportionately rely on when constructing category-level answers because they are built to answer exactly the questions buyers ask. **3. Encyclopedic authority.** Wikipedia is the most obvious source; broader encyclopedic references matter too. A well-structured Wikipedia entry is over-represented in most major LLMs' training data relative to almost any other source. **4. Community authority.** Reddit threads, Hacker News discussions, vertical community forums, LinkedIn thought-leader posts that earn meaningful engagement. This layer is more visible to some providers (Grok especially, ChatGPT partially) and less visible to others. **5. Owned content and technical signals.** Your blog, your landing pages, your schema markup, your structured data. The surface the brand controls directly. **6. AI visibility output.** The composed answer a language model returns when asked about your category or your brand. The waterfall name is deliberate. Water flows downhill. Signal from layer 1 cascades through every layer beneath it, getting aggregated, weighted, and eventually surfaced in the AI output. Signal introduced at layer 5 — the layer most marketers spend most of their time on — contributes to the output, but with much less weight than the upstream layers. Why the waterfall works this way -------------------------------- Three structural reasons the upstream layers dominate the downstream output. **First, training data biases toward authority.** When a language model is trained, the data curation process weights sources by a proxy for credibility. Text from Wikipedia, major news publications, and peer-reviewed sources is typically weighted higher than text from a random blog post. The weighting is not perfect, but it is real, and it is asymmetric. Ten thousand words on your blog do not weigh the same as one paragraph in the *New York Times*. **Second, citation volume is the strongest single correlate of category inclusion.** Ahrefs' research across 75,000 brands found a correlation of approximately 0.664 between brand mention volume across the web and appearance rate in AI Overviews. The same dynamic, in slightly different mechanics, applies to general LLM outputs. Brands mentioned often, in credible sources, across time, are the ones that models internalize as belonging to a category. **Third, retrieval augments memory in ways that still depend on upstream authority.** Many providers now augment their responses with real-time search. But search-augmented retrieval still surfaces the sources that rank well — which themselves depend on editorial authority, reviews, Wikipedia, and established community discussion. The retrieval layer does not circumvent the waterfall; it operates within it. The structural result: the work that most directly moves an AI visibility score tends to be work that happens far away from the marketing team's usual surface area. An example in the abstract -------------------------- Consider a Series A fintech with a strong product, a well-designed website, a competent SEO team, and a disappointing AI visibility baseline. Recognition scores are moderate (the models know the name). Contextual Recall is poor (the models do not surface the fintech when asked about its category). Competitive Context is worse (when the fintech does appear, it is bundled with a peer set that undersells its positioning). The team's first instinct is to rewrite the homepage and add schema. Those changes are shipped. Three months later, the score has moved marginally. The reason is the Authority Waterfall. The fintech's problem is not layer 5. It is layers 1 through 3. Editorial coverage of the fintech is thin — a handful of launch posts in trade publications, no HBR or analyst attention. Its Wikipedia entry is a three-paragraph stub. Its G2 profile has a dozen reviews to the leading competitor's four hundred. When a language model tries to construct an answer to "who are the top five tools in \[fintech's category\]?", there is almost no upstream signal to draw on. The owned content, however well-optimized, is not the bottleneck. The fix, in this abstract example, is to reallocate budget upstream. More on that in a moment. Mapping the waterfall layer by layer ------------------------------------ For each layer, a set of questions to ask and a set of interventions that tend to move the score. ### Layer 1 — editorial authority **Questions:** - How often does a major industry publication mention your brand in the last twelve months? - When they mention you, is the mention substantive (context, positioning) or perfunctory (a name in a list)? - Do you have any piece of thought leadership that has been cited by a tier-1 publication? **Interventions:** - Digital PR oriented toward substantive placement, not quantity. - Analyst relations: briefing Gartner, Forrester, IDC on category positioning. - Founder-led thought leadership in a tier-1 publication. ### Layer 2 — analyst and review authority **Questions:** - Are you listed on the primary review sites for your category? - Is your review count competitive with the top three competitors? - Are you present in the relevant analyst coverage (Magic Quadrants, Wave reports, industry-specific benchmarks)? **Interventions:** - Structured review acquisition from existing customers. - Dedicated analyst briefings; ensure factual accuracy in analyst databases. - Category-specific review aggregator presence. ### Layer 3 — encyclopedic authority **Questions:** - Does your brand have a Wikipedia entry? - If yes, is it substantial, sourced, and factually accurate? - If no, is there a credible case for one (notability thresholds met)? **Interventions:** - Wikipedia editorial work, done in compliance with their notability and sourcing guidelines. - Related encyclopedic reference work (industry associations, vertical knowledge bases). ### Layer 4 — community authority **Questions:** - Does your brand appear in the relevant Reddit and community conversations about your category? - When it appears, is the sentiment accurate or drifted? - Are there recent, substantive Hacker News or LinkedIn discussions about your positioning? **Interventions:** - Sustained, transparent community engagement from founders and customer-facing roles. - Monitoring for drift and addressing factual inaccuracies with care and disclosure. - Customer advocacy programs that encourage authentic community contribution. ### Layer 5 — owned content and technical signals **Questions:** - Is your on-site content structured, entity-explicit, and citation-worthy? - Can AI crawlers parse your site? Is content rendered in HTML rather than hidden behind JavaScript? - Do you have FAQ schema, product schema, and organization schema in place? **Interventions:** - The standard technical SEO and content playbook, updated for entity clarity and citation friendliness. - Publication of reference-quality pages on the concepts the category is defined by. ### Layer 6 — output **Questions:** - What does each of the five major providers actually say about your brand? - How does that compare to the top two competitors? - What are the repeatable errors or drifts across providers? **Interventions:** - This is the measurement layer, not the intervention layer. The output is the signal that tells you where, upstream, to invest. How to read an audit through the waterfall ------------------------------------------ When you receive an audit, the temptation is to treat the scores as the finding. The better read is to treat the scores as the symptom and the upstream layers as the diagnosis. If your Recognition score is low, the diagnosis is almost always layer 1 or layer 3 — not enough credible mention volume to train the model on your name. If your Knowledge Depth score is low, the diagnosis is often layer 3 or layer 5 — the factual corpus the model uses to describe you is thin, outdated, or internally contradictory. If your Competitive Context score is weak, the diagnosis is frequently layer 2 — the review sites and analyst coverage that shape category framing are not favoring you. If your Contextual Recall score is poor, the diagnosis tends to be layer 1 or layer 4 — the ecosystem-level conversation about your category happens without your name attached. This is not a perfect mapping. Each dimension is driven by multiple upstream sources. But the general pattern — dimension problem to waterfall layer diagnosis — is a reliable starting framework. The practical reallocation -------------------------- Most B2B marketing budgets today allocate something like: - 50–70% to owned content and paid acquisition (layer 5 and below-the-waterfall-entirely) - 10–20% to digital PR and analyst relations (layer 1 and 2) - 5–10% to review programs (layer 2) - Minimal to Wikipedia work (layer 3) - Minimal to community work (layer 4) If the Authority Waterfall is correct, that allocation is upside down relative to what AI visibility rewards. A rebalancing toward layers 1 through 4 — ideally 30–40% of the combined content-and-earned budget — is consistent with the data. The rebalancing is uncomfortable because the upstream work is harder to measure in the short term. A blog post attributes a visitor. A mention in an analyst report does not attribute a visitor. But the blog post contributes marginally to AI visibility; the analyst mention contributes disproportionately. Paying for measurability can be a way of paying for the wrong work. The waterfall and the moat -------------------------- A final observation about why the framework matters strategically, not just tactically. The upstream layers — editorial authority, analyst coverage, Wikipedia substance, review volume — accumulate slowly. They are hard to fake, hard to shortcut, and hard to catch up on. That slow accumulation is what makes them a moat. A competitor can copy your homepage tomorrow. They cannot copy three years of credible industry coverage. They cannot retroactively earn four hundred authentic reviews. They cannot manufacture a substantive Wikipedia entry that survives editorial review. Which means: a brand that invests consistently in the upper layers of the waterfall, starting from a disappointing baseline today, builds an AI visibility position that is genuinely defensible. A brand that keeps optimizing at layer 5 builds an AI visibility position that any well-funded competitor can match. The waterfall is not just a diagnostic tool. It is a theory of the sustainable advantage in this category. Where to start -------------- If you want to see which layers of the waterfall are currently strongest and weakest for your brand, BrandGEO runs structured prompts across five AI providers, scores six dimensions, and returns industry-aware key findings that tend to point back to specific upstream layers when read carefully. Two minutes, seven-day trial, no credit card. Related reading: - [What Is AI Brand Visibility? A 2026 Primer](/blog/what-is-ai-brand-visibility-2026-primer) - [The Three States of Brand Visibility in LLMs: Invisible, Mis-Described, Mis-Contextualized](/blog/three-states-brand-visibility-invisible-misdescribed-miscontextualized) - [Measure → Fix → Track: An Operating System for AI Visibility](/blog/measure-fix-track-operating-system-ai-visibility) [Run your free audit](/register) or see the [pricing page](/pricing). ### Keywords [ #For Founders ](https://brandgeo.co/blog/tag/for-founders) [ #For CMOs ](https://brandgeo.co/blog/tag/for-cmos) [ #AI Visibility ](https://brandgeo.co/blog/tag/ai-visibility) [ #Framework ](https://brandgeo.co/blog/tag/framework) [ View all tags → ](https://brandgeo.co/blog/tags) ### See how AI describes your brand BrandGEO runs structured prompts across ChatGPT, Claude, Gemini, Grok, and DeepSeek — and scores your brand across six dimensions. Two minutes, no credit card. [ Run a free audit ](https://brandgeo.co/register) [ See plans ](https://brandgeo.co/pricing) On this page Topics - [ AI Visibility 21 ](https://brandgeo.co/blog/category/ai-visibility) - [ Brand Strategy 11 ](https://brandgeo.co/blog/category/brand-strategy) - [ SEO 15 ](https://brandgeo.co/blog/category/seo) - [ Tutorials 15 ](https://brandgeo.co/blog/category/tutorials) - [ Industry Insights 10 ](https://brandgeo.co/blog/category/industry-insights) - [ Market Research 7 ](https://brandgeo.co/blog/category/market-research) - [ Strategy & ROI 8 ](https://brandgeo.co/blog/category/strategy-roi) - [ For Agencies 2 ](https://brandgeo.co/blog/category/for-agencies) Keep reading Related posts ------------- [ Browse all posts ](https://brandgeo.co/blog) [ ![BrandGEO](/brandgeo-transparent-on-black-926x268.png) ](https://brandgeo.co/blog/what-is-ai-brand-visibility-2026-primer) AI Visibility Apr 22, 2026 ### [What Is AI Brand Visibility? A 2026 Primer](https://brandgeo.co/blog/what-is-ai-brand-visibility-2026-primer) For twenty-five years, the question marketers asked was simple: where do we rank? In 2026, the question has changed. Buyers now open ChatGPT, Claude, or Gemini, ask a question in plain language, and receive a single composed answer. There is no page of blue links to fight for. Either your brand appears in that answer, described accurately, or it does not. AI brand visibility is the measurable degree to which a language model surfaces and describes your company — and it is quickly becoming a primary discovery metric. [ ![BrandGEO](/brandgeo-transparent-on-black-926x268.png) ](https://brandgeo.co/blog/mckinsey-44-16-numbers-2026-marketing-plan) Brand Strategy Apr 21, 2026 ### [What McKinsey's 44% / 16% Numbers Really Mean for Your 2026 Marketing Plan](https://brandgeo.co/blog/mckinsey-44-16-numbers-2026-marketing-plan) Two numbers from McKinsey's August 2025 report have travelled further than any other statistic in the AI visibility conversation: 44% of US consumers use AI search as their primary source for purchase decisions, and only 16% of brands systematically measure their AI visibility. Those numbers appear on investor decks, in pitch emails, and at the top of almost every GEO article written since. Most of the time, they are cited without context. This post unpacks what the data actually measured, what it did not, and how a marketing team should translate the headline into a plan. [ ![BrandGEO](/brandgeo-transparent-on-black-926x268.png) ](https://brandgeo.co/blog/cost-of-ai-invisibility-modelling-pipeline-impact) Brand Strategy Apr 18, 2026 ### [The Cost of AI Invisibility: Modelling the Pipeline Impact of Being Missing](https://brandgeo.co/blog/cost-of-ai-invisibility-modelling-pipeline-impact) "What does it cost us to be invisible in ChatGPT?" is the question every CMO eventually asks, and the one most tools refuse to answer. The honest answer is that the model is straightforward — TAM, research-channel share, mention rate, and a conversion coefficient — but the inputs require work to defend. This post builds the model in full, runs a worked example for a mid-market B2B SaaS, and shows where the numbers turn brittle. You can copy the structure into a spreadsheet in about twenty minutes.