What Is Share of Voice in AI Search?

Share of Voice (SoV) in AI search is the percentage of AI-generated answers in your category that mention your brand. If you track 100 prompts like "best CRM for small businesses" across ChatGPT, Perplexity, and Gemini, and your brand appears in 35 of the responses, your AI Share of Voice is 35%. It answers the fundamental question: when consumers ask AI about your category, how often does your brand show up?

This is the metric that separates brands with genuine AI visibility from brands that appear once or twice and assume they're covered. A single mention in one ChatGPT response is anecdotal. A 40% Share of Voice across 75 prompts and 5 platforms — that's a defensible position.

How AI SoV Differs from Traditional Share of Voice

Traditional Share of Voice has existed in marketing for decades. In media, it measures your brand's share of total advertising impressions. In PR, it measures your share of media coverage. In social, it measures your share of online conversation. AI SoV measures something fundamentally different.

The most important distinction: traditional SoV is primarily a function of spend. AI SoV is primarily a function of content quality, source authority, and technical accessibility. You can't buy your way to a higher AI Share of Voice — though OpenAI's early moves into ChatGPT advertising may change this over time.

Why AI SoV Is More Meaningful than Position Rank

In traditional SEO, position rank is the core metric — are you #1, #3, or #10 for a keyword? In AI search, position rank matters less because LLMs don't produce ranked lists. They synthesize information into conversational answers where multiple brands may appear in a single response, or only one brand may be mentioned.

Share of Voice captures the right question: across the full range of prompts a consumer might ask about your category, how consistently does your brand appear? A brand ranking #3 in a single AI response but appearing in 50% of all category responses has stronger AI visibility than a brand ranking #1 in a single response but appearing in only 10% of them.

How to Measure AI SoV

AI Share of Voice cannot be tracked with a pixel, a tag, or a Google Search Console query. Because LLMs are non-deterministic — the same prompt run five times returns five different responses — measurement requires multiple runs per prompt, aggregated over time.

The formula: AI SoV = (your brand mentions / total brand mentions across tracked prompts) × 100.

What you need:

1. A prompt library — a set of non-branded, category-relevant prompts that represent how consumers search your category (e.g., "best luggage for frequent travellers," "affordable CRM for startups India")
2. Multi-platform coverage — run each prompt across ChatGPT, Perplexity, Gemini, Claude, and Grok
3. Multiple runs — run each prompt multiple times (Cited uses 5 runs per prompt per day) to account for response variance
4. Aggregation — average mention rates across runs, prompts, and platforms to produce a stable SoV number

Manual approach: Run 20-30 category prompts across 3 platforms. Log which brands appear in each response. Calculate mention percentages. This gives you a directional SoV estimate but won't account for response variance or platform differences at scale.

Automated approach: Platforms like Cited automate this across 5 AI platforms with daily refresh cycles, tracking up to 125 prompts per category with multi-run aggregation.

What Drives AI SoV

Five factors determine whether your brand's Share of Voice grows or shrinks:

1. Prompt coverage. Are you visible across a broad range of category prompts, or only a narrow subset? Brands with high SoV appear on "best X" prompts, "X vs Y" comparison prompts, "how to choose X" prompts, and category-specific intent prompts. Brands with low SoV appear only on prompts that directly match their positioning.

2. Cross-platform consistency. A brand with 60% SoV on Perplexity but 5% on ChatGPT has a platform concentration problem. Cited Index data shows platform variance of up to 96 percentage points for the same brand — 96% mention rate on one platform, 0% on another. True AI visibility requires presence across all five major platforms.

3. Third-party authority. Since 95.7% of AI citations come from third-party sources, your SoV is largely determined by how well your brand is represented on review sites, publications, Reddit, and comparison pages — not just your own website.

4. Content structure. AI platforms cite content that directly answers questions with specific data. Brands whose content leads with answers, includes comparison tables, and uses structured data see higher mention rates.

5. Recency. AI platforms with real-time web search (Perplexity, Gemini) reward fresh content. Brands that publish regularly and update key pages maintain higher SoV than brands with static websites.

AI SoV in the Cited Index

The Cited Index tracks AI Share of Voice across 202 Indian brands in 8 categories — the largest public AI visibility benchmark for the India market. Each brand's AI Visibility Score incorporates SoV alongside mention rate, sentiment, and platform coverage, measured across ChatGPT, Gemini, Claude, and Perplexity.

The methodology page details the exact scoring formula, prompt archetypes, and aggregation approach. SoV is calculated at the category level — a brand's share is measured against all tracked competitors in its category, not against the entire index.