Share of voice used to be simple. You added up your owned, earned, and paid presence across the channels your buyers used — Google, LinkedIn, podcast ads, trade press — divided by the total category footprint, and got a percentage. Higher number, more mindshare.
That math still works. It’s just no longer enough.
Buyers now skip the channels SOV was built to measure. When a head of growth at a 200-person SaaS asks “what’s the best customer data platform under $5K/month?”, she isn’t typing it into Google. She’s asking ChatGPT. Or Perplexity. Or the AI Overview Google now puts above the blue links. She gets back a paragraph that names two or three vendors. Those vendors are the consideration set.
If you weren’t named, you weren’t outranked. You were never on the list.
That’s AI share of voice — the percentage of category-relevant AI answers in which your brand gets mentioned. It’s a new metric, but it maps onto an old marketing fundamental: when buyers ask the gatekeeper for a recommendation, what fraction of the time do they hear your name?
This post is the practitioner’s guide to measuring it, understanding what drives it, and starting to grow it this quarter.
Why AI Share of Voice Matters Now
Three things happened at once.
Buyer behavior shifted. Gartner, Forrester, and every B2B agency tracking it agree: AI assistants are now a primary research channel for technical buyers. The exact share varies by category, but the trajectory is the same — the percentage of pre-vendor-call research happening inside an AI chat is climbing fast and isn’t reverting.
The funnel got shorter. A traditional SOV channel (a podcast ad, a sponsored post, a paid search term) is one impression in a long sequence. An AI answer is the whole top of funnel compressed into 60 words. There’s no page two. No “also consider.” If you’re not in the answer, you don’t exist for that query.
Most brands have no instrumentation. GA4 won’t tell you whether ChatGPT mentioned you yesterday. Your SEO platform won’t either. The dashboard you’ve been measuring marketing performance on is genuinely silent on this channel. Brands that don’t measure AI share of voice today aren’t being out-marketed — they’re being out-instrumented.
The compounding kicker: AI models trained on the current web crystallize today’s answers into tomorrow’s defaults. Every quarter you’re invisible, you’re letting the model harden its mental model of your category around competitors’ names.
How to Measure AI Share of Voice
You don’t need a six-figure platform to start. The manual method is good enough to build the muscle and produces directionally correct numbers.
Step 1: Build a seed query list
Pick 10–20 queries that real buyers ask in your category. Three sources to pull from:
- Sales calls. What questions do prospects ask before they pick a shortlist? “What are the best [category] tools for [use case]?” “Who competes with [incumbent]?” “What’s the alternative to [legacy vendor]?”
- Search Console. The non-branded queries that bring people to your site are the same queries they’re now typing into ChatGPT.
- Reviews and forums. Reddit threads, G2 questions, Slack community posts — these are the language buyers actually use.
Aim for a mix: broad (“best [category] tool”), use-case-specific (“[category] tool for [vertical]”), comparison (“alternatives to [competitor]”), and decision-stage (“how to choose a [category] platform”).
Step 2: Run each query, log what comes back
Run every seed query through at least three assistants — ChatGPT (with browsing), Perplexity, and Google’s AI Overview. Add Claude and Gemini if you have time. For each response, record:
- Which brands were named
- What position they appeared in (first mention is meaningfully more valuable than third)
- Whether your brand was a mention (named in the prose) or a citation (linked source the model pulled from)
- What sources the AI cited (Reddit, G2, your own site, a tier-1 publication)
Mention vs. citation matters. A mention says the model knows you exist for this category. A citation says the model trusts a specific page enough to source from it. You want both.
Step 3: Calculate share of voice
For each query: count brand mentions in the answer. Your AI SOV for that query is (your mentions / total mentions) × 100. Average across the 10–20 queries to get your category-level number.
A reasonable benchmark: under 5% means you’re effectively absent. 10–20% is a foothold. 30%+ is dominance — you’re one of the two or three brands the AI defaults to.
Run the same set monthly. The delta is the signal — what you’re doing is working when the line goes up across queries. For a deeper breakdown of related metrics, see how to measure AI search visibility.
What Determines Your AI Share of Voice
Once you start logging answers, a pattern emerges fast: the same brands keep getting cited, even when they aren’t the biggest in the category. Four signals do most of the work.
Third-party citations and reviews. AI models reward sources that aren’t you. G2, Capterra, TrustRadius, “best of” roundups, analyst coverage, podcast transcripts, Wikipedia. The number of credible third-party sources mentioning your brand in your category context is the single highest-leverage input.
Recency of coverage. Live-retrieval models (Perplexity, AI Overviews, ChatGPT browsing) weight fresh content. A 2025 review beats a 2022 review. Brands that publish, get covered, and refresh content within the last 6 months show up disproportionately.
Answer-ready, structured content. Models lift content they can extract cleanly. Tight definitions in the first 50 words under a heading. Comparison tables. Numbered steps. Schema markup (Organization, Product, FAQPage, Review). If your content reads like a sales pitch, the model skips it for a competitor’s clean definition.
Brand entity consistency. If your homepage says “customer data platform,” your G2 listing says “CDP,” and a guest post calls you a “marketing analytics tool,” the model can’t form a coherent association with your category. Brands that win SOV use the same category language and one-line positioning everywhere they appear.
How to Grow Your AI Share of Voice
The work is unglamorous, compounding, and durable.
Get reviewed where the models look. G2, Capterra, TrustRadius, Product Hunt, Trustpilot if you’re DTC. Aim for 30+ recent reviews per major directory. The recency signal matters as much as the volume.
Earn third-party media coverage. Pitch tier-2 industry publications with original data — proprietary benchmarks, survey results, pricing studies. They publish; you get cited. Five well-placed pieces in the next quarter beat fifty backlinks from low-authority blogs.
Publish thought leadership your category cites. Take a real position. Make a contrarian claim and back it with data. Brands that are quoted by other people in the category get pulled into AI training data far more often than brands that publish only on their own site.
Write GEO-optimized content. The same questions buyers ask AI are the headings on your blog. Answer the question in the first 50 words. Add a comparison table where it fits. Use schema markup. (Deep dive: how to write content AI search engines actually cite.)
Get on podcasts where transcripts are published. Models read transcripts. Four to six episodes in the next six months, on shows your category listens to, is a high-leverage move that almost no SaaS brand systematically pursues.
Align messaging across every surface. Audit how your brand is described on your homepage, About page, G2, Capterra, LinkedIn, Crunchbase, Wikipedia, every guest post. Pick one category phrase. Pick one one-line positioning statement. Rewrite anything inconsistent.
The 10-Minute AI Share of Voice Audit
Run this today. You’ll have a defensible baseline by the end of your next coffee.
- Pick three seed queries. Best [your category] tool. [Your category] for [your top vertical]. Alternatives to [your top competitor]. Two minutes.
- Run each in ChatGPT, Perplexity, and Google AI Overview. Nine total responses. Three minutes.
- Log brand mentions in a spreadsheet. Columns: query, assistant, brand mentioned, position, mention vs. citation. Three minutes.
- Calculate three numbers. Your mention count. Total brand mentions across all responses. Your AI SOV percentage. Two minutes.
Whatever number you get is your starting point. If your brand wasn’t named once across nine responses, that’s not a measurement problem — that’s the problem you actually need to solve.
Where to Go From Here
The brands winning AI search in 2026 are the ones who started measuring in 2025. There’s a real first-mover dynamic: every quarter you compound third-party mentions, schema, and answer-ready content is a quarter your competitors aren’t. Models trained on the current web bake today’s mention frequency into tomorrow’s defaults.
The 10-minute audit above gives you the baseline. The next step — actually closing the gap — is where most teams stall, because it requires honest visibility into which signals you’re missing across every assistant, every query, and every competitor.
That’s what we do at Luminari. A free professional AI visibility audit runs your brand through ChatGPT, Perplexity, Gemini, and Claude across the queries that actually matter in your category. We surface your AI share of voice, the citation gaps, and the specific signals to fix first.
Find out where you stand today.
We’ll run your brand through ChatGPT, Perplexity, Gemini, and Claude across the queries that matter in your category, and show you exactly which signals are missing.