SonarLens Blog
The AI Impression You Can't Buy (Or Even See)
Millions of people ask ChatGPT and Gemini for product recommendations every day. Whether your brand shows up — and how it's described — isn't determined by your ad spend. Here's what's actually going on, and what you can do about it.
Every day, millions of people open ChatGPT or Gemini and ask questions like:
- "What's the best CRM for a growing startup?"
- "Which running shoes should I get for marathon training?"
- "What skincare brand is worth trying?"
And then the AI answers. Confidently. With specific recommendations.
Here's the uncomfortable truth: you have almost no control over what it says about your brand — and until recently, you couldn't even see it.
The invisible brand channel
For the past two decades, brand managers have obsessed over search rankings, social media share of voice, and review scores. These channels are measurable, manageable, and contestable. Outspend your competitor on Google, and you can outrank them. Buy better reviews, and you can (in theory) game the system.
AI recommendations don't work like that.
When ChatGPT recommends a brand, it's drawing on a complex mix of training data, web content, user conversations, and sources it has learned to trust. There's no ad slot. No SEO trick that translates directly. No amount of spend that guarantees you'll be recommended.
What matters is your AI brand presence: how your brand is discussed across the sources AI learns from, what associations it has built up over time, and how it compares to competitors in the eyes of the model.
And until very recently, that presence was completely invisible.
Why this matters right now
The scale is already significant. OpenAI reported over 500 million weekly active users in early 2025. Google's Gemini is integrated into Search, Android, and Workspace — reaching billions of touchpoints daily.
More importantly: these aren't passive impressions. When someone asks an AI which brand to choose, they tend to act on the answer. AI recommendations carry the credibility of an expert friend — personalized, conversational, apparently unbiased.
The category leader in AI recommendations may not be the one spending most on advertising. It's the one whose story has been told most coherently across the sources AI trusts.
What AI actually knows about your brand
AI models build their understanding of a brand from:
Content your brand controls (website, blog, press releases) — but filtered through the model's interpretation of how authoritative and relevant it is.
Third-party coverage (reviews, editorial, forums, Reddit threads, industry publications) — often weighted more heavily than owned content.
Competitive context (how your brand compares to alternatives mentioned in the same conversations and articles).
Temporal patterns (newer sources often have more influence, but foundational associations from years of coverage are hard to shift).
The result is a brand image that's partially a product of everything you've ever put out — and partially beyond your control.
The demographic gap nobody talks about
Here's where it gets more interesting: AI doesn't give everyone the same answer.
ChatGPT and Gemini tailor their recommendations based on the user's context — what they've discussed before, where they say they are, what budget they've indicated. A question about running shoes gets different answers for a competitive marathoner in Germany versus a casual jogger in New Zealand.
This means your brand might be well-recommended to one audience segment and barely mentioned to another — without you knowing either.
Understanding your AI recommendation rate across demographics is exactly the kind of segmented insight that used to require expensive research panels. It now requires asking the right questions of the AI systems themselves.
What good AI brand intelligence looks like
The brands that are getting ahead of this problem are doing a few things:
1. Measuring, not guessing. Rather than assuming AI recommends them well, they're running systematic studies — asking the same questions AI users would ask, across multiple models, across their key demographics.
2. Benchmarking against competitors. The question isn't just "does AI mention us?" — it's "does AI mention us more or less than our competitors, and why?"
3. Understanding the sources AI trusts. The citations in AI recommendations reveal which websites, publications, and forums are shaping brand perception. Those sources become content and PR targets.
4. Tracking over time. AI brand presence isn't static. It shifts as new content is indexed, as competitors act, as models update. Periodic measurement catches meaningful changes early.
5. Connecting AI visibility to business outcomes. As AI becomes a more significant discovery channel, brands are starting to treat AI recommendation rate as a leading indicator alongside traditional brand health metrics.
The window to act is now
Most brands have not yet taken AI brand visibility seriously as a measurement and strategy problem. That's a window — but not a permanent one.
The brands building AI brand intelligence practices today will have a meaningful advantage as AI's role in consumer discovery continues to grow. They'll know what the models say about them. They'll know why. And they'll know what to do about it.
The ones still ignoring it will find out the hard way — when they discover their competitors are being recommended and they're not.
SonarLens runs structured studies across ChatGPT and Gemini, using realistic consumer profiles to show you exactly what AI recommends in your category — and why. Reports ready in minutes, no subscription required.