Perception Is the Product: The New Blind Spot in AI Search And Why Marketers Must Go Old School Again
The New Blind Spot
If you’ve read the recent Search Engine Land article claiming that “nearly all ChatGPT users still visit Google,” you probably felt that mix of curiosity and skepticism we all get when a bold data point hits the feed. I did too. Because here’s the thing, trusting any article like that requires more than just glancing at a chart. It means asking: Who’s behind the data? What are they really measuring? And what aren’t they showing us?
The truth behind the article, Search Engine Land, is a reputable industry source. They’re not some random blog spinning AI hype. The data they cited came from SimilarWeb, showing that 95.3% of ChatGPT users also visited Google in the same month. That’s a valid observation, but not the full story. SimilarWeb’s data can’t tell us why those users visited Google, what they searched for, or how often they switched between platforms. And since the report was based on one month’s snapshot, it’s not clear if the overlap is shrinking, growing, or just normal internet behavior.
Still, the claim fits the narrative many want to believe: “ChatGPT isn’t replacing search.” But here’s the kicker none of us outside OpenAI or Google have access to the real data that could prove it either way.
The data gap that’s shaping perception OpenAI doesn’t share query-level data.
That means no one, not marketers, not analysts, not even journalists; can see what people are asking inside ChatGPT, which prompts succeed or fail, how long sessions last, who’s returning daily versus dropping off, or how usage differs by country or device.
We’re all working from shadows and estimates. OpenAI holds all of that data internally. And it’s not because they’re disorganized, it’s because that information is their moat. It’s their competitive advantage, their privacy shield, and their narrative control all in one. Just like Google never shares full keyword data anymore, OpenAI doesn’t have to. They’ve learned the same lesson: perception is power. If you control the story, you don’t have to prove it with numbers.
The Transparency Paradox
This is the tension at the heart of AI search: massive visibility, minimal transparency. Platforms like OpenAI and Google hold all the behavioral data; the rest of us are left inferring what’s happening through traffic estimates and social noise.
ChatGPT is one of the fastest-growing products in history, yet nobody outside the company can see the real metrics. We can’t measure prompts per session, abandonment rates, or true retention. That vacuum is exactly what drives OpenAI’s branding strategy, and it’s why marketers should pay close attention.
Brand as the Strategy When Data Goes Dark
When you can’t share data, you must shape perception. OpenAI understands that perfectly.
Brand as the moat. Without public metrics to prove superiority, OpenAI built emotional trust instead. “ChatGPT” doesn’t just describe a product; it defines a category. Like “Google it” once did for search, “ChatGPT it” now stands for intelligent assistance.
Perception as momentum. Every keynote, feature release, and celebrity demo fuels the idea that ChatGPT is everywhere, in classrooms, boardrooms, and brainstorms. That perception draws in users, journalists, and investors, creating a self-reinforcing loop where brand awareness becomes user growth.
Narrative over numbers. Instead of dashboards, OpenAI offers a story: progress, safety, and innovation. It’s the same principle marketers use when attribution is fuzzy: control the narrative so people fill in the blanks in your favor.
The branding evidence. Multiple studies and articles back this up. Search Engine Journal reported that 77% of ChatGPT users treat it like a search engine, suggesting that people already perceive it as a replacement or complement to Google. Meanwhile, Search Engine Land found that while ChatGPT use is high, 95% of those users also visit Google, showing how perception leads the reality. Academic research, such as “How ChatGPT Usage Behavior Influences Its Branding in Terms of Consumer-Based Brand Equity” (ResearchGate, 2024), supports this too, noting that usage behavior itself strengthens brand equity.
Skepticism still matters. Not all data support the dominance narrative. Search Engine Land also reported that Google’s overall search volume continues to grow, and that ChatGPT’s share of “search-style” behavior remains below 1%. Studies comparing Google and ChatGPT performance on commercial queries show Google still outperforms AI on accuracy, speed, and depth. Together, these facts suggest two parallel truths: ChatGPT’s brand dominance is real and deliberate, but the data behind its actual impact is still opaque.
Frequently Asked Questions About Marketing in the Age of AI Search
Q1. How can marketers figure out what people are really asking ChatGPT? Short answer: You can’t access query logs, but you can rebuild them manually. Long answer: Recreate your audience’s search behavior yourself. Run ChatGPT sessions as if you were your own customer, note phrasing patterns, and log which questions produce relevant or confusing answers. Interview clients about how they’ve used AI for research and watch Search Console for “how-to” keywords disappearing, signs that those questions have migrated into ChatGPT.
Q2. How can we measure when AI fails to deliver good answers? Short answer: Test it like a customer would. Long answer: Conduct “mystery-shopper” tests. Ask ChatGPT questions your prospects would then grade responses for helpfulness, bias, and accuracy. Follow up with real prospects: “Did you try using AI before contacting us?” Patterns reveal where human insight still beats automation.
Q3. Is ChatGPT use growing, or are people getting tired of it? Short answer: Both trends exist; it depends on who you ask. Long answer: Because OpenAI doesn’t publish retention data, you have to look for social and behavioral signals. Build a small user panel to journal AI use for 30 days, and monitor community chatter for “AI fatigue.” Even your team’s usage curve can mirror broader sentiment.
Q4. Where and how are people using AI most? Short answer: It varies by device, context, and confidence. Long answer: Prompt from multiple devices and VPN locations to detect localization differences. Compare geographic data in your analytics, then survey your audience. A few hundred responses can reveal regional adoption trends that global reports can’t show.
Q5. Which ChatGPT features actually matter to users? Short answer: The ones that solve real friction points. Long answer: You can’t see OpenAI’s internal usage stats, but you can infer trends by tracking YouTube tutorials, LinkedIn conversations, and what clients mention unprompted (“I used ChatGPT to write this…”). Add a question to your lead forms, “Did you use AI before reaching out?” to collect real behavioral insight.
Q6. Is ChatGPT really replacing Google? Short answer: Not yet, but it’s changing the journey. Long answer: Studies show heavy overlap between ChatGPT and Google users, which means people toggle between both. Run dual experiments: optimize content for SEO and AEO (Answer Engine Optimization). Ask new leads what they did first, search or prompt, and track where your brand surfaces in AI screenshots or conversations.
The Dual Reality
| OpenAI’s View | Marketers’ Reality |
|---|---|
| Perception equals market share | Data gaps create uncertainty |
| Brand substitutes for transparency | Attribution gets harder |
| Control the story | Rebuild direct observation |
| Hype drives adoption | Context drives trust |
Both realities are true. OpenAI’s brand dominance shows what’s possible when a company markets through perception rather than proof. But it also reminds us why human marketers still matter most: we can see what the dashboards can’t.
The Takeaway: Re-Humanizing Marketing
AI has made the digital landscape measurable but mysterious. We’re drowning in data yet blind to behavior. The marketers who win next won’t be the ones chasing algorithms they’ll be the ones watching people. If you can’t measure it, you can still understand it. If the data disappears into a black box, go back to curiosity, conversation, and observation.
Because sometimes, the most powerful marketing strategy isn’t the one that shows every number it’s the one that makes people believe in what you stand for.
