What Is AI Share of Voice? The Brand Mention Share Metric You’re Probably Missing

Showing up in AI answers and winning in AI answers are not the same thing, most brands don’t know the difference, and their dashboards aren’t telling them.

I ran an AI Brand Visibility campaign for Lululemon in Similarweb and found that the brand’s visibility looked strong on paper. A significant share of all AI answers in the activewear space mentioned Lululemon.

But when I looked at brand mention share (AKA AI share of voice), the metric that tells you how much of the competitive conversation you actually own, the picture changed. Lululemon was showing up in AI answers, yes. But those same answers were packed with competitors.

That gap between visibility and mention share is the story that brand visibility score alone doesn’t tell. In this guide, I’ll break down exactly what AI share of voice is, how it’s calculated, why it differs from brand visibility, and what you can actually do with it, using the Lululemon campaign data throughout to show what it looks like in practice.

What is AI share of voice & how is it calculated?

AI share of voice (AI SOV) is the percentage of all brand mentions across AI-generated responses that belong to your brand. It measures how much of the competitive conversation your brand owns relative to every other brand named in the same answers.

The formula is straightforward:

In Similarweb’s AI Brand Visibility tool, this metric is called Brand Mention Share and lives at the top of the Brand Overview report.

It’s worth being precise about what “total brand mentions” means here, because this is where methodology matters, and where a lot of AI SOV tools quietly mislead you.

Some tools calculate your share of voice using a closed denominator: they only count mentions of the brands you’ve manually added as competitors. If you’re tracking five brands, your SOV is your mentions divided by the sum of those five brands’ mentions.

The problem is obvious once you see it. If a brand you didn’t think to add is getting mentioned constantly in your category, say, a fast-growing challenger you weren’t watching, it’s invisible in the denominator, and your SOV looks artificially strong.

Similarweb counts every brand that appears across tracked AI responses, whether you listed them or not. So when I ran the Lululemon campaign, the 11,295 total brand mentions included every brand AI engines actually named: Athleta, Nike, CRZ Yoga, Vuori, Gymshark, but also Reddit, YouTube, and others that showed up organically in the responses. Lululemon’s 11.87% reflects its share of that real, unfiltered competitive landscape, not a curated list that flatters the result.

How does AI Share of Voice differ from brand visibility percentage?

This is the question I get most often from marketers new to AI visibility tracking, and I think the confusion is understandable because both metrics seem to measure the same thing. They’re not.

Brand visibility % answers: Are you in the room? It measures the share of AI-generated answers in which your brand appears at all. Lululemon at 26.72% means the brand showed up in 1,341 of the 5,018 answers the campaign tracked.

AI Share of Voice answers: How much of the room’s attention do you hold? It measures your brand’s mentions as a proportion of all brand mentions across those same answers. Lululemon at 11.87% means that, across those 5,018 answers, Lululemon accounted for roughly one in nine mentions.

Here’s a simple way to think about the difference. Imagine AI answers as dinner party conversations. Brand visibility tells you how many conversations you showed up in. AI Share of Voice tells you how much of the talking you did compared to everyone else at the table.

The Lululemon data makes this concrete:

The gap between those two numbers isn’t a problem, it’s information. It tells you that Lululemon’s AI answers are competitive, crowded affairs with multiple brands named in each response. That’s a different strategic situation from a brand with, say, 10% visibility and 8% mention share, where it dominates the responses it appears in.

Both metrics matter. Visibility without mention share context is incomplete. Mention share without visibility context is misleading. You need both to understand where you actually stand.

What is a good AI Share of Voice percentage?

I’ll be honest: there’s no universal answer, and anyone who gives you a single benchmark number without category context is oversimplifying.

In Yoga Apparel, the brand’s home category, the one it built its reputation in, Lululemon holds 30.63% mention share. That’s a dominant position in a competitive field that includes Athleta, CRZ Yoga, and Vuori. In Joggers & Sweatpants, the brand holds 5.08% mention share and only recently moved to the top spot on the leaderboard.

Both numbers are meaningful. Neither is objectively good or bad without the competitive context.

What I’d suggest instead of chasing a specific percentage is tracking three things based on our data:

Your mention share trend over time

A rising mention share, even from a low base, tells you your content and citation strategy is working. The rank flip Similarweb detected, Lululemon moving from #2 to #1 in Joggers & Sweatpants, is exactly this kind of signal. Something shifted in the AI conversation for that topic, and it’s worth understanding what drove it.

Your mention share versus your top competitors

In the competitor LLMs comparison I looked at, Lululemon’s average visibility over the past 7 days was 27%, more than double Nike’s 10% and Athleta’s 10%, and well ahead of CRZ Yoga, Vuori, and Gymshark at 5% each. That gap is a competitive moat worth protecting, not just a number to report.

Your mention share across topics

The 25-point spread between Lululemon’s Yoga Apparel mention share (30.63%) and its Joggers & Sweatpants mention share (5.08%) tells you exactly where to focus content and citation efforts. High visibility, low mention share in a topic means competitors are constantly named alongside you. Low visibility and improving mention share mean you’re gaining ground in a category you haven’t fully penetrated yet.

How do you improve your AI share of voice?

After looking at the Lululemon data, the improvement roadmap becomes fairly clear. Here’s how I’d think about it:

Start with the topics where the gap is largest

Lululemon’s Shorts for Training topic shows 12.40% visibility but only 6.68% mention share, with Nike, Gymshark, Under Armour, and Ten Thousand all competing in the same answers. That means AI engines are consistently naming multiple competitors alongside Lululemon in those responses.

The question to ask is: for which specific prompts is Lululemon absent, and what content exists from competitors that earns those mentions? Similarweb’s Prompt Tracking tool shows exactly this: the individual prompts firing in each topic and which brands appear in the responses. It’s one of the most useful GEO tools available for closing prompt-level gaps before competitors widen them.

Create answer-first content for the prompts where you’re missing

A 2024 Research by Princeton University found that content structured with direct answers and cited statistics achieves significantly higher visibility in AI-generated responses.

AI engines retrieve content that leads with a direct, clear answer to the question. Pages structured with a strong opening answer, not a preamble, not a brand story, are the ones that get pulled into responses.

For Lululemon, a page that opens with a direct answer to “what are the best shorts for HIIT training” is more extractable than one that opens with Lululemon’s brand heritage.

Build third-party brand mentions in the right places

AI models weigh brands referenced in independent, authoritative sources more heavily than those that appear only in self-published content. The Topics Summary in the Lululemon campaign shows Reddit and YouTube appearing as top source categories alongside e-commerce brands across several topics, which suggests where AI engines are sourcing their brand knowledge. Community presence, press coverage, and editorial mentions in relevant publications all feed the citation footprint that drives AI SOV growth.

Watch the rank flip signals

The insight Similarweb surfaced- that Lululemon rose from #2 to #1 in the Joggers & Sweatpants topic, is exactly the kind of signal that tells you when a strategy is working. These daily leaderboard shifts occur due to content changes, new third-party mentions, or competitor activity. Tracking them gives you a feedback loop that’s much faster than waiting for quarterly reviews.

Track weekly, not quarterly

AI model behavior shifts constantly. A brand can lose meaningful share of mention within weeks if a competitor publishes strong content or earns prominent third-party coverage. The 7-day window in the Lululemon campaign already shows trend lines moving, which means that visibility and mention share data have directional value precisely because it’s recent.

The Bottom Line

AI share of voice is not a vanity metric. It’s the number that tells you whether your brand is genuinely winning the AI conversation or just showing up.

Brand visibility gets you in the room. AI share of voice tells you how much of the room you actually own. And as AI engines become the default research tool for more buyers across more categories, the brands that track both and act on the gap between them are the ones that will build durable competitive positions in this channel.

The Lululemon data I walked through in this guide is a real example of what that gap looks like in practice: strong overall visibility, dominant mention share in core topics like Yoga Apparel, and clear room to grow in contested categories like Shorts for Training and Joggers & Sweatpants. That kind of granular, topic-level picture is what makes AI share of voice actionable, not just a number to report, but a roadmap for where to focus next.

If you want to run the same analysis for your brand, Similarweb’s AI Search Intelligence tracks brand-mention share, brand visibility, prompts, citation sources, and sentiment across ChatGPT, Gemini, Perplexity, and Google AI Mode, all in a single campaign view.

FAQ

What is the difference between AI share of voice and brand visibility in AI?

Brand visibility percentage measures whether your brand appears in an AI-generated response at all. AI share of voice measures what percentage of all brand mentions across those responses belong to your brand. Visibility tells you if you’re present, mention share tells you how competitive that presence actually is.

What is the difference between AI share of voice and citation share?

AI share of voice counts how often your brand name is mentioned relative to other brands across AI-generated responses. Citation share measures how often your content URLs are linked or referenced as a source by those same AI engines. A brand can have strong share of voice but low citation share, the model mentions the brand without linking to its content, or the reverse.

How often should I track my AI share of voice?

Track AI share of voice in a weekly basis at minimum, daily if you’re in an active competitive category. AI model behavior and competitor content moves can change your mention share within days, not months.

Can a small brand build meaningful AI share of voice?

Yes. AI share of voice reflects content coverage, AI topical authority, and citation footprint, not raw brand size or ad spend. A smaller brand that publishes answer-first content on its core topics and earns citations from relevant third-party sources can outperform a larger brand that hasn’t optimized for AI visibility.

Why is my AI share of voice low even though my brand visibility is high?

This is actually very common, and it’s not necessarily a red flag. High visibility with low mention share means your brand is appearing in AI answers, but so are a lot of competitors in those same answers. It tells you the category is competitive and crowded rather than that your brand is underperforming. The next step is to look at the topic level, find the specific topics where your mention share is weakest relative to competitors, and focus your content and citation efforts there first.

Is AI share of voice the same across different industries?

Not at all, and this is one of the most important things to understand before benchmarking your numbers. In highly fragmented categories with dozens of competing brands, like activewear or beauty, even the market leader might hold 10–15% mention share because AI engines spread their recommendations widely. In more concentrated categories with a few dominant players, the top brand might own 40–50% of all mentions. This is why comparing your AI share of voice to a brand in a different industry is meaningless. The only benchmarks that matter are the ones inside your own category, tracked consistently over time.

by Maayan Zohar Basteker

Senior SEO Specialist at Similarweb

Maayan is a senior SEO specialist with 7+ years of experience in SEO. She loves complex research projects, creating SEO strategies and performing technical audits.