I was scrolling through my analytics last week and noticed something weird. Traffic dipped, but the people who did show up were buying faster. Turns out a ton of them came from ChatGPT, not Google. That tracks with what this savvy professional shared in a recent walkthrough: visitors arriving from AI tools convert at roughly 3x the rate of traditional search clicks.
Most teams still optimize for blue links. The creator behind this video flips the script and shows how to optimize for AI answers instead, using HubSpot’s new AEO tool to peek inside the black box of ChatGPT, Gemini, and Perplexity.
The old way vs. the new way
Here’s the contrast that stuck with me:
- SEO (old way): Rank on Google. Chase clicks. Measure positions and impressions. Win by keywords.
- AEO (new way): Get mentioned inside AI answers. Measure citations, share of voice, and sentiment. Win by being the source AI pulls from.
The author calls AEO “answer engine optimization,” and the point is simple. When someone asks ChatGPT “what’s the best AI tool comparison site,” you either show up in that answer or you don’t. There’s no page two to fall back on.
What the HubSpot AEO tool actually tracks
The expert walked through the dashboard live, and four numbers stood out:
- 📊 Brand visibility score: how often your brand appears across tracked prompts. His Futurepedia brand scored 66%.
- 💬 Brand sentiment: scored from -100 (very negative) to positive, showing how AI describes you.
- 🎯 Share of voice: your brand mentions vs. competitors across the same prompts.
- 📎 Citations breakdown: which sources AI pulls from (blogs, guides, listicles) and whether they’re earned or owned.
The prompts tracked aren’t scraped from the wild. The tool generates them based on your product, ICP, and buyer journey stages (awareness, consideration, decision). You can edit, delete, or add prompts that matter to your business.
How to set it up (the contributor’s flow)
The original poster ran through the free trial in real time. Here’s the path he took:
- Drop in your brand name and website domain.
- Confirm brand variations and business type.
- Review the auto-detected competitor list, edit if needed.
- Pick your main product and ideal customer profile (he chose Skill Leap and small business owners).
- Review the 10 suggested prompts. Delete any that miss the mark, add ones tied to real buyer questions.
- Wait a couple minutes for the tool to analyze your visibility across AI engines.
That’s it. The setup took less time than writing a single blog post.
The recommendation engine is where it gets useful
This is the part I think most people will care about. The tool doesn’t just score you and walk away. It tells you exactly what content to make to improve.
In the walkthrough, the AI suggested a listicle blog post with a specific title, summary, target audience, and reasoning. The logic: across the prompts being tracked, listicles were the format AI tools cited most often. So the recommendation matched the format the answer engines actually rewarded.
The person who shared this also pointed out that recommendations come prioritized. Work the high-priority items first, then move down the list.
Practical use cases worth thinking about
A few ways this becomes more than a vanity dashboard:
- Competitive gap analysis: if a competitor dominates share of voice on a key buying-stage prompt, you know exactly where to publish next.
- Content format strategy: stop guessing whether to make a guide, a listicle, or a comparison page. Let the citation breakdown tell you.
- Sentiment monitoring: catch negative framing in AI answers before it spreads. The author noted his sentiment was strongly positive, but another brand he tested scored zero with stacks of issues to fix.
- Buyer-stage targeting: prompts get tagged by awareness, consideration, or decision stage. Spend energy on the stage where you’re weakest.
Challenges to keep in mind
A few honest caveats from watching the demo:
- The free trial only shows ChatGPT data. Gemini and Perplexity unlock at the $50/month tier.
- AI answers shift constantly, so a snapshot today won’t match next month. Treat this as ongoing monitoring, not a one-time audit.
- The prompt list is your strategy. Bad prompts in, useless data out. Spend real time picking questions your buyers actually ask.
- Recommendations point at content gaps, but you still have to write the content well. The tool diagnoses, you execute.
My take
I was genuinely surprised at how concrete the output is. Most “AI visibility” tools I’ve seen are vague dashboards. This one names the blog post you should write next. That’s the kind of clarity busy marketers need.
If AI search is eating your traffic, watch the full walkthrough from the original creator for the complete dashboard tour and a closer look at the citations and recommendations sections.