How to Measure Your Brand’s Share of Voice Inside AI

Once you accept that discovery is happening inside AI tools, an uncomfortable question follows pretty quickly:

How do you even know if your brand is showing up there?

With Google, we had dashboards.

Rankings.

Impressions.

Clicks.

Search Console.

Ahrefs.

SEMrush.

With AI?

It mostly feels like shouting into a fog and hoping for the best.

But measurement is starting to catch up — just not in the way we’re used to.

If you’re not cited, you’re invisible

This is the core shift.

In traditional search, you could rank on page two and still exist. You were at least present.

In AI-driven discovery, there is no page two.

There is:

• mentioned

• or not mentioned

If your brand doesn’t appear in the answer, it’s effectively invisible to that buyer at that moment.

That’s why share of voice inside AI tools is becoming a much more meaningful metric than raw traffic.

Traffic is becoming a lagging indicator

One of the traps teams are falling into is obsessing over what they can still measure easily.

Organic traffic.

Referral traffic.

Direct traffic.

Those numbers matter — but they’re now downstream effects, not leading indicators.

AI influences:

• what people believe

• what they shortlist

• how they frame questions

• which brands feel “obvious”

By the time someone visits your site, a lot of the decision has already been shaped.

Which means visibility before the click matters more than ever.

The HubSpot AEO Grader: a useful starting point

HubSpot has released a free AEO (AI Engine Optimisation) grader, and while it’s not perfect, it’s a very good place to start.

You plug in:

• your brand

• your location

• your industry

• your products or services

And it evaluates how your brand shows up across major LLMs (ChatGPT, Gemini, Perplexity) relative to competitors.

What it’s useful for:

• understanding brand recognition inside AI

• seeing who the models think your competitors are

• getting a directional sense of share of voice

• spotting gaps in how your brand is described

What it’s not:

• real-time

• precise

• a definitive scorecard

Training data lags. Some models are months behind. Changes you make today may only show up much later.

But directionally? It’s extremely helpful.

Why competitor context matters more than your score

The most valuable part of these tools isn’t the number.

It’s the comparison.

AI doesn’t evaluate you in isolation. It always evaluates you relative to alternatives.

Seeing:

• who gets mentioned instead of you

• what they’re known for

• how they’re positioned

is far more actionable than obsessing over whether your score is 62 or 68.

This is also where AI itself becomes useful — take the outputs and ask:

“Where are we strong compared to X?”

“Where are we missing entirely?”

“What themes do competitors own that we don’t?”

Expect messy data (for now)

This part requires a bit of patience.

Analytics hasn’t caught up yet.

In many cases:

• AI-driven visits show up as direct traffic

• or referral traffic

• or disappear entirely

HubSpot is ahead of Google Analytics here. We can already see AI referrals broken out clearly in HubSpot, whereas GA often can’t identify the source yet.

Some platforms (Cloudflare, for example) are starting to surface more of this data, but it’s uneven.

Which means you should expect:

• partial visibility

• inconsistent attribution

• incomplete reporting

That’s not failure — it’s early-stage infrastructure.

Measurement is moving from precision to direction

This is a big mindset shift for performance teams.

We’ve been trained to expect:

• exact numbers

• clear attribution

• deterministic funnels

AI breaks that.

What matters now is:

• trend direction

• relative visibility

• consistency of mentions

• alignment between brand messaging and AI outputs

You’re no longer measuring “how many people clicked this page”.

You’re measuring “how often are we part of the answer”.

Monetisation is coming (it always does)

It’s also worth being realistic.

Right now, we’re in the “golden age” of AI discovery — similar to early Google before ads dominated everything.

That won’t last.

These platforms will monetise.

Sponsored answers will appear.

Paid placement will creep in.

We’ve seen this movie before:

• Google

• LinkedIn

• Meta

• Amazon

Which is exactly why organic presence inside AI tools matters now, while the ground rules are still forming.

What actually moves the needle today

Based on what we’re seeing across clients and our own site, a few things consistently help:

• clean, accurate data

• structured, specific content

• pages that answer one question well

• FAQs that mirror real prompts

• consistency across platforms (site, video, communities)

• removing outdated or misleading information

None of this is flashy.

All of it compounds.

From traffic-first to visibility-first

The hardest shift isn’t technical — it’s mental.

For years, success meant:

“More traffic.”

Now it increasingly means:

“More visibility in the right places.”

Sometimes that leads to a click.

Sometimes it doesn’t.

But it always shapes perception.

And perception is what drives the funnel before the funnel even reaches your website.

The quiet truth about this shift

No one has this fully figured out yet.

Not Google.

Not OpenAI.

Not HubSpot.

Not agencies.

Not clients.

We’re all experimenting in public.

But one thing is already clear:

If your brand isn’t being mentioned inside AI-generated answers, you’re losing ground — even if your dashboards still look “fine”.

Search hasn’t disappeared.

It’s just stopped behaving the way we were taught to expect.

And the sooner you start measuring visibility where discovery actually happens, the better positioned you’ll be for whatever comes next.