AI Visibility Scoring: How to Measure Your Brand's Presence in AI Answers

If you search for "best email marketing platform" on ChatGPT, it generates an answer and cites 3 or 4 sources. If you ask the same question on Perplexity, it cites 6 to 10. On Gemini, another 3 to 5. Every time one of these AI engines answers a query in your industry without citing your brand, you are losing visibility to a competitor who did get cited.

The problem is that traditional SEO metrics do not capture this. You can rank position 2 on Google for a keyword and still be completely absent from AI-generated answers for that same query. Conversely, a page ranking position 7 might get cited by Perplexity because it contains a specific data point that the AI system needed. AI visibility scoring is the metric that fills this gap.

This article explains what AI visibility scoring is, how to calculate it, what benchmarks to aim for, and how to use Serpent API's AI Rank endpoint to automate the tracking process.

What Is AI Visibility Scoring?

An AI visibility score measures how frequently your brand or domain appears in AI-generated search answers across a defined set of keywords. It is the AI equivalent of a traditional SEO visibility index—instead of tracking your position in a list of links, it tracks whether AI systems cite or mention you when answering relevant questions.

Types of AI Mentions

Not all AI mentions are created equal. When evaluating your visibility, it helps to distinguish between three types of appearances:

• Direct citation: The AI links to your page as a source. This is the strongest form of visibility because it includes a clickable link that drives traffic. Example: "According to Mailchimp's 2026 benchmark report [source], the average email open rate is 21.3%."
• Brand mention: The AI names your brand in its response without linking to you directly. This builds awareness but does not drive clicks. Example: "Popular options include Mailchimp, ConvertKit, and Brevo."
• Recommendation: The AI specifically recommends your product or service. This is the highest-value mention because it carries the AI's implicit endorsement. Example: "For small businesses with limited budgets, Mailchimp's free plan is the best starting point."

The Components of a Visibility Score

A comprehensive AI visibility score accounts for four components: citation frequency (how often you appear), citation position (whether you are the first source cited or the last), cross-engine coverage (which AI platforms cite you), and mention context (citation, brand mention, or recommendation).

Why Traditional Rank Tracking Is No Longer Enough

Traditional rank tracking tools monitor your position in Google's organic results. This remains valuable, but it misses a growing portion of how people discover information. Consider this scenario:

Your company ranks position 1 on Google for "best accounting software for startups." Your rank tracker shows green arrows. But when someone asks ChatGPT the same question, your brand is not mentioned at all—it recommends three competitors instead. When someone asks Perplexity, your brand appears as the 6th citation (essentially buried). When Gemini generates an AI Overview for this query on Google itself, your page is not cited even though you rank first in the organic results below it.

In this scenario, your traditional SEO metrics look perfect while your actual visibility to a growing share of searchers is poor. AI visibility scoring captures what rank tracking misses.

How to Calculate Your AI Visibility Score

Basic Score: Citation Rate

The simplest version of an AI visibility score is the citation rate: the percentage of monitored keywords where your domain appears as a citation in at least one AI engine.

Basic AI Visibility Score:
= (Keywords where your domain is cited / Total keywords monitored) x 100

Example:
- You monitor 100 keywords relevant to your business
- Your domain is cited in AI answers for 18 of them
- Basic AI Visibility Score = 18%

Weighted Score: Accounting for Quality

A more sophisticated approach weights citations by quality. A first-position citation on ChatGPT is worth more than a 7th-position citation on Perplexity. A direct recommendation is worth more than a passing mention. Here is a weighted scoring model:

Weighted AI Visibility Score:
For each keyword, calculate points:
- Citation position 1-2:     3 points
- Citation position 3-5:     2 points
- Citation position 6+:      1 point
- Brand mention (no link):   0.5 points
- Recommendation:            +2 bonus points

Engine multipliers (based on search volume):
- ChatGPT:    1.5x
- Gemini:     1.5x
- Perplexity: 1.0x
- Claude:     0.8x

Weighted Score = Sum of (points x engine multiplier) / Max possible score x 100

Per-Engine Breakdown

Always calculate your visibility score per engine in addition to the aggregate. This reveals platform-specific weaknesses. For example, if your score is 25% on Perplexity but only 5% on ChatGPT, you likely have a Bing indexing problem (since ChatGPT uses Bing). If your score is strong on ChatGPT but weak on Gemini, your Google E-E-A-T signals may need work.

AI Engine	Keywords Cited	Visibility Score	Action if Low
ChatGPT	14 / 100	14%	Check Bing indexing and rankings
Gemini	11 / 100	11%	Strengthen E-E-A-T signals
Perplexity	22 / 100	22%	Maintain current strategy
Claude	9 / 100	9%	Publish more primary-source content

Benchmarks: What Is a Good Score?

AI visibility benchmarks vary by industry, competition level, and the keyword set you are monitoring. However, based on analysis across multiple industries, here are general guidelines for core topic keywords (not broad, unrelated queries):

• Below 5%: Poor. Your brand is largely invisible in AI search. Immediate action needed on content strategy and multi-engine indexing.
• 5% to 15%: Average. You have some AI presence but are not a go-to source. Focus on content depth, freshness, and structured data.
• 15% to 25%: Strong. Your brand is regularly cited by AI engines. Maintain and expand content coverage to grow further.
• Above 25%: Exceptional. You are a dominant source in your niche. This level is typically achieved by industry publications, major SaaS platforms, and domains with deep topical authority.

Important context: these benchmarks apply to your core topic keywords—the 50 to 200 keywords most relevant to your business. If you monitor 1,000 broad keywords tangentially related to your industry, expect much lower scores.

Tracking AI Visibility with Serpent API

Serpent API's AI Rank endpoint lets you programmatically query AI search engines and analyze citations. Here is how to build an automated AI visibility tracker.

Step 1: Define Your Keyword Set

Start with 50 to 100 keywords that represent your core business topics. These should be informational queries that AI systems are likely to answer (not navigational queries like brand names). Group them by topic cluster for better analysis.

Step 2: Query AI Engines via Serpent API

const API_KEY = "your_serpent_api_key";
const MY_DOMAIN = "yourdomain.com";
const keywords = [
  "best email marketing platform",
  "email marketing automation tools",
  "email deliverability tips",
  // ... more keywords
];

async function checkAIVisibility(keyword) {
  const res = await fetch(
    `https://apiserpent.com/api/ai-rank?q=${encodeURIComponent(keyword)}&engine=all&apiKey=${API_KEY}`
  );
  const data = await res.json();
  const results = {};

  for (const [engine, result] of Object.entries(data.results || {})) {
    const citations = result.citations || [];
    const myIndex = citations.findIndex(c => c.url?.includes(MY_DOMAIN));
    results[engine] = {
      cited: myIndex >= 0,
      position: myIndex >= 0 ? myIndex + 1 : null,
      totalCitations: citations.length
    };
  }
  return results;
}

// Run weekly and store results
async function weeklyVisibilityCheck() {
  let citedCount = { chatgpt: 0, gemini: 0, perplexity: 0, claude: 0 };

  for (const kw of keywords) {
    const result = await checkAIVisibility(kw);
    for (const engine of Object.keys(citedCount)) {
      if (result[engine]?.cited) citedCount[engine]++;
    }
  }

  const total = keywords.length;
  console.log("AI Visibility Scores:");
  for (const [engine, count] of Object.entries(citedCount)) {
    console.log(`  ${engine}: ${((count / total) * 100).toFixed(1)}%`);
  }
}

Step 3: Track Trends Over Time

Store each weekly check's results in a database or spreadsheet. Plot your visibility score over time for each AI engine. Look for trends over 30 to 90-day windows. A sudden drop across all engines often indicates a content freshness issue. A drop on one specific engine may indicate a technical issue (indexing, robots.txt changes) or an algorithm update.

Step 4: Compare Against Competitors

Run the same keyword set for your top 3 to 5 competitors. This gives you a competitive visibility landscape: who is dominating AI citations in your space, which competitors are growing or declining, and which keywords represent opportunities where no competitor has strong AI visibility.

5 Ways to Improve Your Score

1. Fill Cross-Engine Gaps

If your per-engine analysis reveals that you are strong on one AI engine but weak on another, the fix is often technical rather than content-related. Verify your indexing on the relevant search engine (Google for Gemini, Bing for ChatGPT). Check that you are not blocking any AI crawlers in robots.txt. Submit your sitemap to all major search engine webmaster tools.

2. Add Specific Data to High-Potential Pages

Identify pages that rank well organically but are not being cited by AI engines. The most common issue is that the content is too generic. AI systems prefer pages with specific, quotable data: percentages, dollar amounts, benchmark results, survey findings. Add original data points to your highest-value pages.

3. Restructure for Extractability

Review your top pages through the lens of an AI system trying to extract a specific fact. Are your key conclusions buried in the middle of paragraphs? Are your headings descriptive enough for an AI to match them to a query? Move important conclusions to the beginning of sections and make headings question-like.

4. Increase Content Freshness

Set a quarterly update schedule for your most important content. When you update, add genuinely new information: new statistics, new examples, updated pricing or features. Update the "Last Updated" date visibly and update the dateModified in your schema markup. AI engines will re-evaluate your content on their next crawl.

5. Build Topical Depth

AI systems evaluate authority at the topic level, not just the page level. A domain with 5 articles about email marketing is less authoritative than one with 30 interlinked articles covering every aspect of the subject. Identify gaps in your content coverage and fill them with detailed, data-rich articles that link back to your pillar content.

For a comprehensive approach to AI search optimization, see our complete ASO guide. To understand the specific factors that influence citation selection, read our analysis of how AI search engines choose which sites to cite.