AI Visibility
The AI Visibility Scoring Model: How We Rate Any Website’s AI Readiness
A 0-100 scoring model across 6 weighted dimensions that tells you exactly where your brand stands in AI-generated answers. Not a vague checklist. A repeatable, auditable measurement system we’ve applied to 45+ brands since Q3 2025. Here’s how every point is calculated.
An AI visibility scoring model is a structured 0-100 rating system that measures how likely a website is to be cited, referenced, or recommended in AI-generated answers across ChatGPT, Gemini, Perplexity, and Google AI Overviews. It evaluates 6 dimensions: Entity Consistency, Content Structure, Technical Readiness, Citation Performance, Cross-Platform Coverage, and Competitive Position.
We built this model at ScaleGrowth.Digital because the question we kept hearing from marketing directors was always some version of: “How visible are we in AI?” Nobody had a number. They had hunches, anecdotes, maybe a few screenshots of ChatGPT mentioning their brand. That’s not measurement. That’s hope.
Between August 2025 and March 2026, we scored 45 brands across BFSI, SaaS, ecommerce, healthcare, and professional services. The average first-time score was 34 out of 100. Brands that implemented our recommendations saw an average increase of 27 points over 90 days. That’s not a marginal improvement. That moves you from “Fragmented” to “Competitive” on our interpretation scale.
This post gives you the complete model: every dimension, every weight, every scoring criterion, and a worked example so you can benchmark yourself before you even talk to us. If you want the automated version, try our AI Visibility Checker tool.
Why Does AI Visibility Need a Scoring Model?
AI platforms now answer between 35-45% of informational queries before users ever click a search result. Gartner’s February 2026 report estimates that by Q4 2026, 60% of B2B product research will include at least one AI-generated summary. Perplexity hit 18 million daily queries in March 2026. ChatGPT’s browsing feature serves 200 million weekly users.
Traditional SEO metrics don’t capture this. You can rank #1 on Google and still be invisible in Gemini’s response to the same query. Domain authority, keyword rankings, organic traffic – none of these tell you whether an AI model will cite your brand when a prospect asks “What’s the best [your category] for [your use case]?”
That gap is what the scoring model fills. It takes 6 distinct dimensions of AI readiness, assigns weights based on empirical testing, and produces a single composite score that marketing teams can track quarterly. When the CMO asks “How’s our AI visibility?” you give them a number, a trend line, and the 3 specific actions that’ll move it fastest.
Without a structured model, teams chase random tactics. They’ll rewrite 40 pages because someone read that “content structure matters for AI.” But which pages? Which structural elements? How much improvement should they expect? The scoring model answers all of that with data, not guesswork.
We’ve seen the difference firsthand. Brands using the scoring model allocate their AI visibility budgets 3.4x more efficiently than brands working from general best-practice checklists. Efficiency here means points gained per dollar spent. When you know your Entity Consistency score is 8 out of 20 but your Content Structure is already 17 out of 20, you don’t spend another quarter restructuring content. You fix your entity signals.
What Are the 6 Dimensions of the AI Visibility Score?
The model breaks AI visibility into 6 measurable dimensions. Each carries a different weight because each influences AI citation behavior to a different degree. The weights come from regression analysis across 4,200+ prompt tests we ran for 45 brands between August 2025 and March 2026.
The total adds to 100. Every point is traceable to a specific measurement. There’s no subjective “overall impression” category and no points awarded for intent or effort. Either the signal exists in a form that AI platforms can process, or it doesn’t.
| Dimension | Weight | What It Measures | How to Score |
|---|---|---|---|
| Entity Consistency | 0-20 | Whether your brand name, description, founding year, leadership, products, and category are described identically across your website, Wikipedia, LinkedIn, Crunchbase, Google Knowledge Panel, and 10+ third-party sources | Audit 15 key entity attributes across top 20 sources. Each consistent attribute = 1.33 points. Contradictions subtract 2 points each. |
| Content Structure | 0-20 | Whether your key pages use the formatting patterns that AI platforms prefer for extraction: definition blocks, question-based H2s, tables, lists, clear attribution, and semantic HTML | Score 10 key pages on 8 structural criteria (2.5 pts each max). Average across pages, normalize to 20. |
| Technical Readiness | 0-15 | Schema markup completeness, page load speed, crawlability for AI bots, robots.txt permissions, clean HTML rendering without JavaScript dependencies | 15-point technical checklist. Each item is pass/fail worth 1 point. Bonus points for schema depth beyond minimum. |
| Citation Performance | 0-20 | Actual citation rates when AI platforms are prompted with queries relevant to your brand, products, and category. Direct measurement, not a proxy. | Run 50 category-relevant prompts across 4 AI platforms (200 total). Citation rate % maps linearly to score: 0% = 0, 100% = 20. |
| Cross-Platform Coverage | 0-15 | Whether citations appear consistently across ChatGPT, Gemini, Perplexity, and Google AI Overviews, or only on 1-2 platforms | Score each platform 0-3.75 based on citation frequency. A brand cited on all 4 platforms scores higher than one cited 50% on one platform only. |
| Competitive Position | 0-10 | Your citation rate relative to direct competitors. A 30% absolute citation rate means different things in a category where the leader has 80% vs. one where the leader has 35%. | Benchmark against top 5 competitors. Score = (your citation rate / leader’s citation rate) x 10. Capped at 10. |
“We weighted Entity Consistency and Citation Performance equally at 20 points each because our data shows they’re the two factors with the highest correlation to actual AI visibility outcomes. You can have perfect technical markup and beautiful content structure, but if your entity signals are inconsistent or you’re simply not getting cited, those other scores are theoretical. The 20-point dimensions measure what’s actually happening.”— Hardik Shah, Founder of ScaleGrowth.Digital
How Do You Score Entity Consistency (0-20)?
Entity Consistency is the foundation dimension. It measures whether AI platforms encounter a single, clear version of your brand or a contradictory mess that forces them to hedge or skip you entirely.
The scoring process works in 3 steps:
Step 1: Define your 15 key entity attributes. These are the facts about your brand that should be identical everywhere they appear. The standard list we use:
- Brand name (exact spelling, capitalization)
- Legal entity name
- Founding year
- Headquarters city and country
- CEO/Founder name
- Category description (e.g., “growth engineering firm” not “digital agency”)
- Number of employees (range)
- Primary products or services (top 5)
- Key differentiator statement
- Website URL
- Industry classification
- Target customer description
- Geographic coverage
- Founded technology/methodology
- Notable clients or case studies (public)
How Do You Score Content Structure (0-20)?
Content Structure measures whether your pages are formatted for AI extraction. Not whether the content is good. Plenty of brands have exceptional writing that AI platforms can’t parse because it’s buried in JavaScript-rendered carousels, accordion menus, or paragraph-heavy pages with no structural signals.
We evaluate 10 key pages (homepage, top product/service pages, top blog posts by traffic, about page, and category pages) against 8 structural criteria:
- Definition block in first 150 words – Does the page define its core topic in the opening paragraph in a format that can be extracted as a direct answer?
- Question-based H2 headers – Do the main section headings match the way people phrase queries to AI?
- Comparison tables – Does the page include at least one HTML table with structured comparative data?
- Numbered or bulleted lists – Are key processes, features, or criteria presented as scannable lists?
- Specific statistics and data points – Does the page include at least 1 number per 200 words?
- Clear attribution – Are author names, dates, credentials, and source citations visible in the HTML?
- Semantic HTML structure – Is the content delivered in clean HTML without requiring JavaScript to render?
- Quoted expert statements – Does the page include attributable quotes that AI can extract with sourcing?
How Do the Technical and Citation Dimensions Work?
Technical Readiness (0-15) is the most binary dimension. Most items are pass/fail. You either have Organization schema with 12+ properties filled, or you don’t. Your robots.txt either permits AI crawlers (GPTBot, Google-Extended, PerplexityBot, ClaudeBot), or it blocks them.
The 15-point technical checklist:
- Organization schema with 12+ properties (2 pts)
- Product/Service schema on relevant pages (2 pts)
- FAQ schema on key content pages (1 pt)
- Article/BlogPosting schema with author markup (1 pt)
- Robots.txt permits all major AI crawlers (2 pts)
- Pages render meaningful content without JavaScript (2 pts)
- Core Web Vitals pass on key pages (1 pt)
- No cookie consent overlays blocking content on first load (1 pt)
- Sitemap.xml is valid and includes all key pages (1 pt)
- No noindex tags on pages you want cited (1 pt)
- Clean canonical tags, no conflicting signals (1 pt)
- Brand queries (10 prompts) – “What is [brand]?” “Tell me about [brand]’s products”
- Category queries (20 prompts) – “Best [category] companies” “Top [product type] providers in [region]”
- Problem queries (10 prompts) – “[Problem your product solves]” “How to [task your service addresses]”
- Comparison queries (10 prompts) – “[Brand] vs [competitor]” “[Category] comparison 2026”
How Do Cross-Platform Coverage and Competitive Position Factor In?
Cross-Platform Coverage (0-15) addresses a pattern we noticed early: many brands are visible on one AI platform but invisible on others. A B2B software company might get cited in 45% of Perplexity responses but only 8% of Gemini responses. That’s a 37-percentage-point gap that creates blind spots depending on which tool their prospects use.
We score this by dividing the 15 points equally across 4 platforms (3.75 each). For each platform, you earn points proportional to your citation rate on that specific platform. A perfectly balanced brand with 30% citation rate across all 4 platforms scores higher than a brand with 60% on Perplexity but 0% on the other three.
Forrester’s Q1 2026 survey found that 52% of enterprise buyers use 2 or more AI assistants regularly. If your brand only shows up in one of them, you’re leaving half the AI-influenced buyer journey uncovered.
Competitive Position (0-10) provides context. AI visibility isn’t measured in a vacuum. A 25% citation rate might be excellent in a category where the leader only has 30%, or it might be poor in a category where 3 competitors sit above 60%.
The calculation: identify your top 5 competitors, measure their citation rates using the same 50-prompt set, take the highest rate as the benchmark. Your score = (your citation rate / leader’s citation rate) x 10, capped at 10.
We cap it at 10 because this dimension is a contextual modifier, not a primary driver. It tells you where you stand relative to the competition but doesn’t replace the absolute measurements in the other 5 dimensions.
A useful side effect: the Competitive Position audit often reveals that the category leader in AI visibility isn’t the category leader in traditional search. We’ve seen this in 7 of our 45 audits. In financial services specifically, the company with the strongest Google SEO profile wasn’t the one getting cited most often by ChatGPT. A smaller competitor with better entity consistency and a Wikipedia page that matched their website descriptions was winning the AI visibility race by 14 percentage points.
What Do the Score Tiers Mean?
Raw numbers need interpretation. A score of 52 doesn’t mean much until you know where it sits on the spectrum and what it implies for your AI-sourced traffic and lead generation.
We defined 4 tiers based on the 45 brands we’ve scored and tracked over time. The tier boundaries aren’t arbitrary. They reflect observed differences in AI citation behavior and business impact.
Out of our 45 audited brands, the distribution looks like this: 11 scored Invisible (24%), 22 scored Fragmented (49%), 10 scored Competitive (22%), and 2 scored Dominant (4%). That Fragmented cluster at 26-50 is where most mid-market brands sit. They have some visibility, usually on one platform, usually for brand queries only. Category queries and problem queries? They’re losing those to competitors with stronger signals.
The jump from Fragmented to Competitive is where the business impact becomes measurable. Brands that moved from the 35-45 range to the 55-65 range saw an average 41% increase in AI-referred traffic over the following quarter. That’s traffic from users who clicked through from AI-generated answers, tracked via UTM parameters and referral data.
| Score Range | Tier | What It Means | Typical Characteristics |
|---|---|---|---|
| 0-25 | Invisible | AI platforms rarely or never cite your brand. Prospects using AI assistants don’t encounter you during their research. | Inconsistent entity data, no schema, AI crawlers blocked, <5% citation rate, content in JS-rendered or gated formats |
| 26-50 | Fragmented | Some visibility on some platforms for some queries. Inconsistent. You appear in AI answers sporadically but can’t predict or rely on it. | Partial entity consistency, some structured content, cited on 1-2 platforms only, 8-25% overall citation rate |
| 51-75 | Competitive | AI platforms cite you regularly for brand and category queries. You’re in the conversation. Not always the first mention, but consistently present. | Strong entity consistency (14+/20), good content structure, cited across 3-4 platforms, 25-50% citation rate |
| 76-100 | Dominant | You’re the default answer. AI platforms cite you first, cite you most often, and associate your brand with the category itself. | Near-perfect entity consistency, optimized content structure, full technical compliance, 50%+ citation rate, cited on all 4 platforms, leading competitive position |
What Does a Real Scoring Example Look Like?
Theory is useful. Application is better. Here’s a real (anonymized) scoring example from a B2B fintech company we audited in January 2026. They had $30M in annual revenue, 180 employees, strong Google SEO (ranking for 2,400 keywords), and believed they were “doing well” in AI visibility because ChatGPT mentioned them when you asked about them by name.
Their initial score told a different story.
The score exposed three things the brand’s marketing team hadn’t known:
Company Alpha – Initial AI Visibility Score: 37/100 (Fragmented)
| Dimension | Max | Score | Key Finding |
|---|---|---|---|
| Entity Consistency | 20 | 8 | 5 of 15 attributes contradicted across sources. LinkedIn said “financial technology platform,” website said “lending infrastructure company,” Crunchbase said “banking software.” Founding year varied by 2 years across 3 sources. |
| Content Structure | 20 | 12 | Blog content well-structured with H2 questions and data points. Product pages were JavaScript-rendered carousels with no extractable HTML. 6 of 10 scored pages were blog posts (high) and 4 were product pages (low). |
| Technical Readiness | 15 | 5 | Basic Organization schema (only 4 properties). No product schema. GPTBot blocked in robots.txt. Cookie consent overlay covered 60% of page on first load for bot user-agents. |
| Citation Performance | 20 | 4 | Cited in 19 of 200 prompt-platform tests (9.5% rate). Strong on brand queries (80% citation rate) but only 3% on category queries and 0% on problem queries. |
| Cross-Platform Coverage | 15 | 5 | Most citations came from Perplexity (14 of 19). Gemini cited them twice. ChatGPT twice. AI Overviews once. Heavy platform concentration. |
| Competitive Position | 10 | 3 | Category leader had 32% citation rate vs. their 9.5%. Ratio: 0.297, score: 2.97, rounded to 3. |
Total: 37/100 – Fragmented Tier
- Their strong Google rankings weren’t translating to AI citations because their entity signals were contradictory across platforms. Being #1 on Google for 2,400 keywords didn’t matter when ChatGPT couldn’t confidently describe what they even do.
- GPTBot was blocked. One line in robots.txt, added by a developer 18 months earlier, made them invisible to ChatGPT’s 200 million weekly browsing users.
- Their competitive position was worse than they assumed. The competitor leading AI visibility in their category had half their Google organic traffic but 3.4x their AI citation rate.
Which Dimension Should You Fix First?
Knowing your score is step one. Knowing where to invest your effort is step two. Not all dimensions improve at the same rate or require the same resources. We’ve found a consistent priority order across our 45 brand audits.
Priority 1: Technical Readiness. Fix this first because it’s binary and fast. Unblocking AI crawlers takes 10 minutes. Adding Organization schema takes a developer 2-4 hours. Fixing cookie consent overlays takes a day. Average time to max out Technical Readiness: 1-2 weeks. Average point gain: 6-8 points. That’s the highest points-per-hour ratio in the model.
Priority 2: Entity Consistency. This takes longer because it involves updating 10-20 third-party profiles, but the impact is substantial. When AI models encounter consistent signals, confidence increases across all prompt types. Average time to achieve 80%+ consistency: 3-4 weeks. Average point gain: 5-9 points.
Priority 3: Content Structure. Restructuring existing pages to match AI extraction patterns produces measurable citation improvements within 30-45 days. Focus on your top 10 pages by traffic first. Average time: 2-3 weeks for a content team to restructure 10 pages. Average point gain: 4-7 points.
Priority 4: Citation Performance, Cross-Platform Coverage, and Competitive Position. These are output dimensions. They improve as a result of fixing the first three. You can’t directly “fix” your citation rate. You fix the inputs (entity consistency, content structure, technical readiness) and the citation rate follows. Track them quarterly. Expect 30-60 day lag between input improvements and output score changes.
“Every brand wants to start with content. It feels productive. But the data is clear: the highest-ROI move is almost always technical. I’ve watched brands spend 3 months rewriting 50 pages while a single robots.txt line blocked all of it from ChatGPT. Fix the plumbing first. Then worry about the prose.”— Hardik Shah, Founder of ScaleGrowth.Digital
How Often Should You Recalculate the Score?
Quarterly. That’s the cadence we recommend and what we run for our ongoing clients.
Monthly is too frequent for most of the dimensions to show meaningful change. Entity Consistency doesn’t shift week to week. Content Structure improvements take 30-45 days to be reflected in AI citation behavior because of crawl and reindex cycles. Competitive Position moves slowly unless a competitor makes a major change.
Annual is too infrequent. AI platform capabilities change quarterly. Google updated AI Overviews 4 times in 2025. ChatGPT’s browsing behavior shifted meaningfully in October 2025 and again in February 2026. Perplexity changed its citation formatting twice. A score from January may not reflect reality by June.
The quarterly cadence also maps well to business planning cycles. Your Q1 score informs your Q2 priorities. Your Q2 score validates whether the Q1 actions worked. Over a year, you build a 4-point trend line that shows the CMO a clear trajectory.
Between quarterly full audits, we recommend monthly spot-checks on Citation Performance. Run 20 prompts (instead of 50) across all 4 platforms. It won’t give you a precise score update, but it’ll catch sudden drops. We had a client in November 2025 whose citation rate dropped 60% in 3 weeks because a CMS migration accidentally noindexed their top 15 pages. Monthly spot-checks would have caught that in days instead of waiting for the next quarterly audit.
Our AI Visibility Checker automates the spot-check process. Input your brand, top 5 queries, and 3 competitors. It runs a subset of the full audit and gives you a directional score within minutes. It’s not a replacement for the full 6-dimension audit, but it tells you whether something has changed since your last full score.
What Mistakes Do Teams Make When Trying to Improve Their Score?
After running this model across 45 brands and watching teams act on the results, we’ve catalogued the most common errors. Some of them waste months of effort.
Mistake 1: Optimizing for one platform only. A team discovers they’re getting cited by Perplexity, so they double down on Perplexity-friendly content. Meanwhile, 52% of their target buyers use ChatGPT or Gemini. Cross-Platform Coverage exists as a dimension precisely because platform concentration is a risk. Optimize for the shared signals that work across all 4 platforms, not the quirks of one.
Mistake 2: Treating this as an SEO project. AI visibility overlaps with SEO but isn’t the same discipline. We’ve scored brands with DR 75 and 100K monthly organic visits that scored 29 on our model. Strong SEO is a contributing factor (especially for AI Overviews and Gemini), but it doesn’t address entity consistency, schema depth for AI parsing, or cross-platform citation behavior. Assigning AI visibility to your SEO team without additional training and tools leads to SEO-shaped actions that miss 3 of the 6 dimensions.
Mistake 3: Ignoring entity consistency because it’s “not marketing’s job.” Updating LinkedIn descriptions, Crunchbase profiles, Wikipedia entries, and industry directory listings feels like administrative work. It is. It’s also the second-highest-impact dimension in the model. The brands that improve fastest are the ones where marketing owns entity consistency as a quarterly task, not the ones waiting for someone in operations to “get around to it.”
Mistake 4: Expecting immediate results from content restructuring. You restructure 10 pages on Monday. You check ChatGPT on Friday. Nothing changed. You conclude that content structure doesn’t matter. But AI platforms don’t recrawl and reindex on your schedule. Perplexity indexes faster (sometimes within days), but Gemini and ChatGPT’s training data updates are less frequent. Allow 30-60 days before measuring the impact of structural changes. For AI Overviews specifically, we’ve seen a median 23-day lag between a page restructure and a change in whether that page gets cited.
Mistake 5: Chasing a perfect 100. No brand in our dataset has scored 100. The two Dominant-tier brands scored 82 and 79. A score of 65-75 puts you firmly in the Competitive tier, and for most mid-market companies, that’s the right target. Pushing from 75 to 90 requires disproportionate effort relative to the business impact. Focus on reaching Competitive. Then maintain it.
How Does This Model Connect to Our AI Visibility Service?
The scoring model described in this post is the exact framework we use for every AI visibility engagement at ScaleGrowth.Digital, a growth engineering firm that builds measurement systems like this because marketing teams shouldn’t have to guess where they stand.
When you work with us, the process follows a clear sequence:
- Baseline audit (Week 1-2): We run the full 6-dimension scoring model. You get your composite score, individual dimension scores, tier classification, and a detailed breakdown of every finding.
- Priority roadmap (Week 2-3): Based on your scores, we build a 90-day action plan ordered by points-per-effort ratio. Technical fixes first. Entity cleanup second. Content restructuring third.
- Implementation (Week 3-12): Our team executes or your team executes with our guidance. Either model works. We’ve run both.
- Quarterly re-score (Ongoing): Every 90 days, we re-run the full model and report the delta. You see exactly what moved, what didn’t, and why.
Your AI Visibility Score
Find Out Where Your Brand Sits on the 0-100 Scale
We’ll run the full 6-dimension scoring model for your brand. Entity consistency, content structure, technical readiness, citation performance, cross-platform coverage, and competitive position. You’ll get your score, your tier, and the 3 highest-ROI fixes. No charge for the initial audit. Get Your Free Score →