HubSpot can now track your AI visibility. Here's what to do with that number

HubSpot has just launched AI visibility tracking inside its SEO tools. Now, you can set up prompts, connect your brand, and see a score. This is a percentage indicating how often your brand appears in AI-generated answers from tools like Gemini and ChatGPT.

While it’s a genuinely useful development, it also surfaces a big problem that many marketing teams currently don’t have a clear answer to. You can now see that you’re invisible, but what you can’t easily see is why or what you can do about it.

Search behavior has already shifted, and the data is catching up

It’s been coming for a while. Buyers are no longer Googling, as they’re asking AI tools for shortlists, comparisons, recommendations, and quick answers. Those same questions your sales teams used to field in discovery calls? They’re now being answered by the likes of ChatGPT before a prospect even gets in touch.

More than 60% of Google searches now end without a single click. Buyers now get the answer they need directly from the SERP or from AI tools entirely, without ever visiting anyone’s website. That means discovery and evaluation are happening somewhere your analytics can’t even see.

Whenever buyers ask AI tools for the best platform for their problem or for leading providers, they instantly get a short list of names. No SERP. No clicking around. No evaluating 10 different pages or results. Just a simple recommendation.

If your brand isn’t in that list, then you don’t exist in that conversion, regardless of your market position or how great your product or service quality is.

This isn’t just in theory, either. In multiple audits we’ve conducted across B2B brands in several sectors, we’ve found companies with genuine expertise and strong reputations that were absent from AI-generated answers in their category. In some cases, their competitors were being recommended, and they had no idea.

That’s because when discovery happens entirely within AI tools, there are no sessions in your analytics to tell you that losses occurred. In these instances, the problem with our clients wasn’t their products but how AI systems failed to interpret their brands.

What HubSpot’s AI visibility tool actually does

To find HubSpot’s new AI feature, you need to:

• Get into your HubSpot portal
• Navigate to Marketing > AEO
• Click Get Started
• Enter your brand name
• Enter your domain
• Click the business type dropdown and select an option
• When finished, click next

To understand how it works, first confirm your brand details. These include your brand name, domain, any variations, and so on. Next, set up a list of prompts, which are questions your buyers might actually ask an AI tool about you. HubSpot runs these prompts daily against supported AI engines and tracks how often your brand appears.

This dashboard will give you:

• Brand visibility score: The percentage of tracked prompts where your brand gets a mention

• Brand visibility over time: So you can see whether things are declining or improving

• Share of voice: How your brand mentions compare to your competitors across the same prompts

• Citation data: Which website AI tools pull from when they respond to your prompts, and whether the sources are yours

It’s a meaningful addition to the toolkit. For example, share of voice across AI answers is something many teams currently do manually, while some haven’t done it at all. However, one thing to remember is that the tool measures only outcomes, not causes. It’ll tell you where you stand, but it won’t tell you why your score is what it is, or what you specifically need to change.

The gap: Measurement is only half of the problem

What honestly happens when most teams get access to an AI visibility score is that they look at it, get alarmed if it’s low, and then are unsure what to do next.

They think they should probably write more content. That first instinct can help at times, but not always, because the reason brands don’t appear in AI answers isn’t just a lack of content but a structural problem.

The part most teams miss is knowing how AI tools like ChatGPT and Gemini decide which brands to recommend. Remember, AI doesn’t just read your website, but it looks for consensus.

Your website is just one source, whereas the LLMs powering AI systems look for agreement across multiple independent sources before they trust a brand enough to cite it. Think LinkedIn content, reliable reviews, industry publications, discussions, news mentions, etc. All of these feed into how AI models understand and categorise your brand.

Research shows that AI looks for signals from 10 or more independent sources before they reliably recommend a brand. Your website accounts for only 5-8% of what influences an AI recommendation, while the remaining 92-95% comes from everywhere else your brand is or isn't talked about.

So, even if you have a well-optimised website and strong SEO, you might still have low AI visibility. Here are the most common causes of low AI visibility we’ve seen across brands:

• AI can’t access your best content: Case studies gated behind forms. Key reports locked behind login pages. Evidence embedded in PDFs. AI crawlers hit these blockers and move on. While the content might be excellent, it’s invisible to the AI systems that would cite it

• Your off-site presence doesn’t build consensus: If your brand isn’t being discussed independently, then AI has nothing more to reference, as one source singing your praises isn’t enough. You need independent validation that AI can find and weigh

• Your brand isn’t clearly associated with your category: AI systems build entity graphs from what they can read across the web. If anything is confusing, such as rebrands, old product names, or subbrands that look like separate companies, then AI tools can produce inconsistent citations even for branded queries

• Trust signals are unreadable or missing: Author attribution, named expertise, structured data that confirms what your business actually does, and third-party validation are the signals that move a brand from being occasionally mentioned to being reliably cited. A lot of brands have some of these, but most don’t have them in a form that AI systems can readily process

• Severe technical issues: Duplicate metadata. Redirect chains. No title tags on pages. Missing schema. No llms.txt files to guide AI crawlers. We see these issues a lot on almost every website we audit. While they don’t stop humans from navigating the website, they do affect how AI systems model the brand

Again, the new HubSpot tool will show you your score. But it won’t identify which of these problems is causing it.

What will move the number?

If you’ve looked at your AI visibility score and want to improve it, here’s how you can do that.

Measure

Start by carefully setting up your prompts in HubSpot. Alongside branded queries, add the category and problem prompts your buyers will use. Track both your citation data and your brand score, so you know which sources AI tools pull from. Here, you want to know where AI finds its consensus and where you aren’t represented.

Diagnose

Understand why the score is what it is. This is the step the tool can’t do for you, and instead, an in-depth AI diagnostic can help. Which prompts are you appearing in? Is your key content accessible? Are your trust signals in a form that AI systems can process? Are there technical issues pressing how AI builds its model of your business?

An AI diagnostic can surface things you and your team don't usually expect, so it’s a step you shouldn’t skip.

Fix

Now, it’s time to prioritise and act. Once you understand the causes, the fixes can be clearer than you might assume. It could be anything from quick wins like ungating key content or metadata completion. It could be to restructure your content with author attribution or convert PDF case studies into crawlable HTML and add internal links.

It could also be strategic to build entity signals and category authority, or even to work on your off-site presence, such as reviews or partner mentions.

Why acting today is critical

Timing is key, and you should focus on this now rather than in six months.

Traditional SEO took years to build an advantage, whereas AI visibility is compounding in months. Brands that establish authority now, while the landscape is still forming, will be the ones AI tools default to citing as the behaviour becomes more mainstream.

Each week a competitor builds their presence is a week they’re harder to knock off the top spot.

Frequently asked questions

How is AEO different from SEO?

Search Engine Optimisation earns rankings; Answer Engine Optimisation improves the signals and structure of pages and content to help AI systems understand, trust and include your content in generated answers.

How many pages does Avidly review in this AEO Audit?

Our AEO Audit starts with a 10‑page representative sample for speed and coverage, but this can be expanded if needed. You'll need to provide us with the 10 most representative URLs, but you can speak to us about how to choose them.

Can Avidly implement the fixes on our behalf?

Yes, we can implement the fixes identified during the AEO Audit. From metadata/schema rewrites to design/CMS updates and any new module work, we’ll scope the implementation as a separate phase.

How often do I need to do an AEO audit of my website?

We advise an in-depth AEO and AI Audit at least once a year, with more frequent light audits performed monthly/quarterly. Implementing the fixes and actions is usually an ongoing task, just like traditional SEO best practices.

The factors that change the urgency/frequency of needing an AEO Audit include that this is a fast-moving and always-evolving field. New product releases, algorithm changes, and learnings on how to be referenced all lead to needing to run audit reports.

How is this different from free audits?

Free AI Audits from tools such as HubSpot and others are definitely useful, but are very surface-level and generic. They're similar to free SEO audits. They only touch on generic best practices or advice and stay at a theoretical level. This is a paid AEO Audit by Avidly, completed by our award-winning team, specific to your URL and goals, and includes a full walkthrough on a call.