Today, more and more shoppers are starting their journey with an AI-powered search. They could ask ChatGPT to compare products or use an AI-powered platform like Perplexity. Or they’re simply Googling an offer and reading the AI overview, all without clicking a link.
HubSpot recognized that our buyers were moving from search engines to response engines like ChatGPT, Gemini, and Perplexity, but we didn’t have a reliable way to measure AI visibility and understand whether our AEO plays were working.
So, in June 2025, the HubSpot Marketing team began working with XFunnel, an AEO tool that allowed us to measure and optimize our AI visibility across ChatGPT, Gemini, Perplexity, and more. Here’s what we learned.
Summary
The results of HubSpot’s AEO strategy
- HubSpot is the #1 CRM 1 more visible
- We saw an 1,850% increase in qualified leads coming from AI
- Leads from AEO convert 3x faster than other sources
- We increased citations by 433%
Building our AEO measurement system
Defining the buyer journey through prompt tracking response engines
The first question we needed to answer was: When a potential customer asks an answer engine about a problem our products solve, is HubSpot in the answer? To find out, we defined the buyer journey through response engines:
Product-driven AEO
We set up XFunnel containers for each product line.
The AEO measurement architecture included:
- Top-level “HubSpot” brand container for overall brand tracking
- Eight dedicated product containers: CRM, Marketing Hub, Sales Hub, Service Hub, Content Hub, Commerce Hub, Data Hub and Breeze
- Feature-specific views within each product container (example: “Email Automation” in Marketing Hub)
This structure allowed subteams to run experiments, track improvements, and optimize the AEO performance of their specific product, while giving us a panoramic view of the entire AEO strategy.
AEO KPIs that we measure
Once the recommendations were defined, we could see and start improving our four main AEO KPIs:
- Response engine visibility (%): how often HubSpot appears for target queries.
- Answer engine voice share (%): how often HubSpot appears for the same queries compared to competitors.
- Response Engine Quotes: how often HubSpot pages are cited as a source in AI responses.
- Response engine citation share (%): how often HubSpot pages are cited for those queries related to our competitors.
How we built our three-pillar AEO strategy
After analyzing the data, we determined that a successful AEO strategy is based on:
- AEO-friendly content on your website with all the information response engines need about your business and its products.
- A strong external presence among the main sources from which response engines are training and extracting information.
We used these foundations to build a three-pillar strategy:
- Optimization of content on the site.
- Off-site amplification.
- Community involvement and forum growth.
Pillar 1: optimization of content on the site
Our AI visibility scores were great from the start, but Xfunnel showed that our citation scores were weak. The response engines did not often refer to HubSpot’s website pages. Brand awareness is the priority, but being mentioned increases the likelihood of influencing response (and generating AI-directed traffic).
After our growth team analyzed Xfunnel’s data, we realized that we needed more ultra-specific content to match the hyper-personalized responses that AI generates for users.
When response engines received purchasing questions or industry fit ratings, they struggled to surface cite-worthy HubSpot content. We needed to create content tailored to our key buyer personas.
“Hubspot will work for My Business?” Personalization comes down to the ability to answer this question.
Industry-specific content
Many potential customers want to understand if a solution is right They industry. We created industry solution pages at scale using an AI content system. We used AI to generate content from HubSpot’s case study library and reviewed it with humans before it was published.
Since we know that AI likes structured data, we have used the breadcrumb and FAQ schema in these industry solutions pages.
92% were cited by response engines, generating a 49% increase in AI visibility.
We have also published software comparison articles for target industries (for example, “5 Best CRMs for Construction Companies”). We saw a 642% increase in citations for those posts and a 58% increase in mentions overall.
Glossary of frequently asked questions about CRM, marketing and sales terms
Our team also found that HubSpot wasn’t appearing enough in the “Problem Exploration” stage of the buyer journey. We launched a Glossary of frequently asked questions covering initial terms like “what is marketing automation?” and “how does lead scoring work?” Each page features a concise definition, related common questions, and links to HubSpot features. Response engines often draw from the content of the definition, and owning these terms means being part of the first response a prospect receives.
As a result, the citation share for related suggestions increased by +60%. Brand visibility for stimuli in the awareness phase increased by +35 percentage points when the glossary was mentioned.
Optimizing product pages for AEO
We’ve updated our product feature pages to better match how answer engines actually understand and retrieve content: adding frequently asked questions, rewriting titles to answer common shopper questions, and improving formatting with tables and lists. We’ve also added structured data to help response engines read and classify pages more easily.
The result: a 56% increase in citations from AI response engines and an improvement in the average ranking position from 1.5 to 1.
Pillar 2: Off-site amplification
Our AEO benchmarks revealed AI responses based on third-party content on HubSpot products. We needed to build HubSpot’s presence in the third-party ecosystem.
Using XFunnel data, our partnerships team identified publishers who have already earned mentions but are not yet mentioning HubSpot. We provided partners with AEO recommendations and templates so they could create response engine-friendly content and get even more citations.
We quickly expanded the program. By the end of 2025, we partnered with hundreds of websites around the world, producing nearly a thousand new pages and garnering hundreds of thousands of new AI citations, all mentioning HubSpot.
Pillar 3: Forum growth
XFunnel benchmarks showed that Reddit was one of the most cited sources for our tracked tips.
Using XFunnel, we’ve integrated always-on Reddit mention monitoring into our reporting, identifying high-impact subreddits and gaps in HubSpot mentions on a weekly basis. We then asked supporters in the HubSpot community to post content that answered some of the top questions buyers have.
After XFunnel benchmarks revealed that the DE and FR markets saw significant growth in mentions on Reddit but zero mentions on HubSpot, we launched localized campaigns. In the space of a month, HubSpot’s mention rate went from 0% to 33.5% in France and 17.1% in DE.
Reddit-driven citations grew from 178 (May 2025) to 146,000 (December 2025).
Everything in this case study started with a single question: When shoppers asked AI, was HubSpot an answer? Before we could optimize anything, we needed to understand which suggestions mattered, where we appeared, who was cited, and where the gaps were.
Measurement came first. The strategy is followed.
This feature is now available in HubSpot. With our AEO tools, you can get real-time visibility into how your brand appears on response engines like ChatGPT, Gemini, and Perplexity.
- Keep track of the instructions your buyers are likely to use
- Compare your visibility with the competition
- Analyze which sources generate citations in response engines
All these insights fuel prioritized recommendations, based on the successful AEO tactics our team has proven, so you know what to improve and start acting on it right away.
And if you use AEO within Marketing Hub Pro or Enterprise, you get timely CRM-based suggestions so your tracking is based on business context from day one, not built from scratch.
