Brand Share of Voice: Tracking It in the Age of AI Search
Luke Costley-White
認知拡大
Expanding recognition
Brand share of voice used to be a social and paid media metric: how much of the conversation in your category belongs to you versus competitors, measured across mentions, impressions, and ad spend. That definition is now incomplete. When someone asks ChatGPT or Perplexity "what's the best tool for X," your share of voice includes whether you get named in the answer. Most share-of-voice platforms, including Meltwater and Brandwatch, still measure social, paid, and organic mentions only. They don't track AI engine citations at all, which means they're reporting an incomplete picture of where your category conversation actually happens now.
What is share of voice, and how do you calculate it?
Share of voice traditionally answers one question: of all the conversation in your market, what percentage mentions your brand? The classic formula divides your brand mentions (or ad spend, or impressions) by the total mentions across you and your named competitors. It's been calculated from social listening data, paid media reporting, and PR coverage for over a decade.
That formula still works. What's changed is where the conversation happens. A growing share of category research now runs through AI engines instead of a Google search box. Someone comparing project management tools, CRM platforms, or marketing systems increasingly asks ChatGPT or Perplexity directly rather than clicking through ten blue links. [DATA NEEDED: current % of B2B buyer research queries routed through AI engines vs. traditional search, from a named 2026 source]. When that happens, the AI engine picks sources, synthesizes an answer, and either names your brand or doesn't. That's a share-of-voice event exactly as real as a mention on X or a segment on the local news, except almost nobody is measuring it that way yet. DOJO AI's Demand Intelligence and AEO monitoring track exactly this: whether your brand gets named when a category question runs through ChatGPT, Perplexity, or Gemini, alongside your traditional search and social share.
Meltwater's own research into AI citations found that a substantial share of what ChatGPT cites comes from news media sources, with that share notably higher on ChatGPT specifically than on other engines [DATA NEEDED: specific Meltwater report name and publication date for this AI-citation source-share finding]. That's a useful data point for a different reason: it shows AI engines have citation patterns and source preferences of their own, separate from Google's ranking signals. A brand can rank on page one of Google and still never get cited by an AI engine, because the two systems are pulling from different evidence.
This is the gap in most share-of-voice tooling. Platforms built for the social-and-PR era measure what they were built to measure: mentions, sentiment, reach, ad share. Bolting on "AI visibility" as an add-on module, if it exists at all, treats AI citations as a side metric rather than a core share-of-voice input. It should be the other way around. If your category conversation is splitting across social, paid, organic, and AI engines, your share-of-voice number needs to be a blend of all four, not three of the four with a caveat.
Traditional share-of-voice tracking vs. AI-engine-inclusive tracking
Traditional SOV tools (Meltwater, Brandwatch, most PR platforms) | AI-engine-inclusive tracking | |
|---|---|---|
Sources measured | Social, paid media, earned press, broadcast | Social, paid media, earned press, broadcast, plus ChatGPT, Perplexity, Gemini, AI Overviews |
Update cadence | Often daily or weekly reporting cycles | Continuous, always-on monitoring |
What counts as a "mention" | A post, article, or ad impression naming the brand | The above, plus a citation or named answer in an AI response |
Competitive benchmarking | Named competitor mention volume | Named competitor mention volume plus which competitor gets cited when the AI answers a category question |
Where it stops | Reports the numbers | Should connect the number to what content, PR, or campaign action closes the gap |
How do you track competitor share of voice without a dedicated analyst?
Share of voice is a comparative metric by definition: your number only means something next to a competitor's. Most teams without a dedicated competitive intelligence analyst either skip this entirely or do it manually once a quarter, scrolling competitor social feeds and Google Alerts before a board deck. That's slow, and it misses the campaigns that matter most: the ones that launch, run for three weeks, and are gone before anyone notices.
A workable substitute for a dedicated analyst is a standing, always-on watch, not a periodic audit. Three things matter here:
Set a fixed competitor list and stick to it. Pick the three to five brands your buyers actually compare you against, not an aspirational list of the biggest names in the category. Share of voice against a competitor set of near-peers means something; share of voice against the market leader ten times your size mostly measures the gap you already know exists.
Track campaign signals, not just brand mentions. A competitor's brand mention count tells you they exist. A spike in their paid social creative, a new landing page pattern, a fresh case study format, or a burst of press coverage tells you they're running a specific push. Watching for pattern changes, not raw mention totals, is what turns monitoring into intelligence.
Automate the sweep so it runs whether or not someone remembers to check. This is the part a solo marketer or a two-person team structurally can't do by hand every week. It's also where an always-on system earns its keep: it can hold a standing watch on competitor content, ad creative, and press mentions and surface the pattern change, instead of waiting for someone to notice on a Tuesday afternoon. DOJO's competitive intelligence guide breaks down the specific signal types worth watching if you're setting this up from scratch, and DOJO AI's Brand Intelligence domain runs this exact sweep continuously rather than on a quarterly cadence.
The output you want from this isn't a mention count. It's an answer to: did a competitor just launch something, and does it change what we should be doing this week?
What does brand intelligence look like without a PR agency?
Brand intelligence used to require a retainer. A PR or brand tracking agency would run quarterly surveys, compile a sentiment report, and present it in a deck a few weeks after the data was actually collected. By the time you saw the numbers, the market had usually moved on from whatever caused them.
Without an agency, the job doesn't disappear, it just needs a different delivery mechanism. Genuine brand intelligence, agency or not, needs four things running continuously rather than quarterly:
Mention capture across every surface that matters, not just the ones a legacy tool happens to cover: social, review sites, forums like Reddit, press, and now AI engine answers.
A comparison point. Sentiment or mention volume in isolation tells you almost nothing. The same numbers next to your two or three closest competitors tell you whether you're gaining or losing ground.
A way to catch AI citation gaps specifically, since this is the surface that's newest and least visible without dedicated tracking. If a competitor gets named when someone asks an AI engine a category question and you don't, that's a real, measurable gap, not a vague feeling.
A route from insight to action. A dashboard that shows sentiment dropped 8 points this month is not brand intelligence. Brand intelligence tells you sentiment dropped because of a specific event, and gives you the next move: a response, a content piece, an outreach.
A dashboard that shows sentiment dropped 8 points this month is not brand intelligence. The tools built to replace the agency retainer are the ones that run all four of these continuously instead of packaging them into a quarterly report. That's a structural difference, not a feature difference: a system that watches every day catches the shift on the day it happens, not eight weeks later in a slide deck. This is the specific gap DOJO AI's Brand Intelligence domain is built to close, connecting mention capture, competitive comparison, and AI citation tracking into one continuously updated view instead of four separate subscriptions.
What is brand intelligence, and which tools track it?
Brand intelligence, at its simplest, is the discipline of knowing how your brand is actually perceived, where that perception is shifting, and why, across every channel where your category gets discussed. It's broader than PR monitoring (which tracks press) and broader than social listening (which tracks social). It includes both, plus reviews, community conversation, competitive positioning, and now AI engine visibility.
We've written the deeper breakdown of what a genuine brand intelligence platform needs to do, and where most vendors stop short, in a separate guide: the marketing intelligence shift. If you're evaluating tools in this category, that's the place to start; this article focuses specifically on the share-of-voice slice of that broader picture.
FAQ
Is brand share of voice the same as brand awareness? No. Awareness measures whether people recognize your brand at all. Share of voice measures your slice of the total category conversation relative to named competitors. A brand can have high awareness and still lose share of voice if competitors are simply talking more, running more paid media, or getting cited more often in AI answers during a given period.
How do I calculate share of voice if I don't have a social listening tool? At minimum, count your brand's mentions across the channels you can access (search, social, press coverage you can find manually) and divide by the total mentions of you plus your named competitor set. It's rough without dedicated tooling, and it won't capture AI engine citations at all, which is the fastest-growing blind spot. It's a starting point, not a substitute for continuous tracking.
Do AI engines like ChatGPT actually count toward share of voice? They should. If a category question run through ChatGPT or Perplexity names your competitor and not you, that's a lost impression in exactly the same way a lost ad auction or an uncredited press mention would be. The difference is that almost no legacy share-of-voice tool is built to catch it.
How often should share of voice be measured? Continuously, if the tooling allows it. Quarterly snapshots miss competitor campaign spikes and AI citation shifts that can happen inside a single news cycle. If continuous monitoring isn't available, monthly is the practical minimum for catching meaningful movement before it's stale.
What's the difference between share of voice and AEO (Answer Engine Optimization)? AEO is the discipline of earning AI engine citations. Share of voice, in its updated 2026 form, is the metric that should include AEO performance alongside social and paid. Think of AEO as one input into a broader share-of-voice number, not a separate scorecard. Our guide to why challenger brands have a 6-12 month window to dominate AEO covers the AEO side in depth, and our comparison of ChatGPT, Perplexity, and Gemini breaks down how citation patterns differ by engine.
Share of voice was never meant to be a vanity number. It was built to answer a practical question: are we gaining or losing ground in the conversation that decides who buyers pick? That question hasn't changed. What's changed is where the conversation happens, and most tracking tools haven't caught up. If your share-of-voice reporting doesn't include AI engine citations yet, it's measuring less of the market than it used to. Talk to DOJO AI about connecting social, paid, and AI-engine share of voice into one number your team can actually act on.
