A quiet but significant shift is underway in how consumers discover products. ChatGPT, the AI chatbot that first captured public attention for writing college essays and generating code, has evolved into something OpenAI’s competitors should have seen coming: a product recommendation engine that millions of people now prefer over Google.
The numbers tell a striking story. According to HubSpot’s research, 55% of consumers have already used ChatGPT for product recommendations. Among those who have tried it, 70% said they preferred ChatGPT’s suggestions over traditional search engine results. Not marginally preferred. Strongly preferred.
That’s a problem for Google. And an enormous opportunity for brands willing to adapt.
From Curiosity to Commerce: Why Consumers Are Abandoning the Search Bar
The dissatisfaction with traditional product search has been simmering for years. Google Shopping results are cluttered with sponsored listings. Amazon reviews feel increasingly unreliable, gamed by sellers running incentivized review programs. Affiliate blogs — once a genuine source of product guidance — have become SEO farms optimized for commissions rather than honest assessment. Consumers know this. They’ve known it for a while. What they lacked was a viable alternative.
ChatGPT filled that vacuum almost by accident. Users who initially came to the platform for writing help or coding assistance started asking casual questions: “What’s the best carry-on suitcase under $200?” or “Which espresso machine should I buy for a small kitchen?” The responses were coherent, personalized, and free of the visual noise that defines modern search results pages. No banner ads. No sponsored placements jockeying for attention. Just answers.
HubSpot’s data reveals the specific reasons consumers cite for preferring AI-generated recommendations. Forty-eight percent said ChatGPT provides more personalized suggestions than search engines. Another 44% valued that the recommendations come with detailed explanations — not just a product name and a link, but reasoning about why a particular item fits the user’s stated needs. The conversational format matters too. Users can follow up, refine their criteria, push back. Try doing that with a Google results page.
There’s also a trust dimension that might surprise skeptics. Despite well-documented issues with AI hallucinations and outdated training data, 46% of respondents told HubSpot they trust ChatGPT’s recommendations as much as or more than those from human experts. That finding deserves a moment of reflection. Nearly half of consumers place an AI chatbot on equal footing with professional reviewers.
Why? Because the perceived neutrality of AI — whether or not it’s fully warranted — stands in sharp contrast to the openly transactional nature of most online product content. Consumers have been trained by years of experience to distrust search results. Every “Top 10 Best” article feels like an ad. ChatGPT, rightly or wrongly, feels like it doesn’t have an agenda.
The demographic breakdown is predictable but still worth noting in its intensity. Gen Z and millennial consumers are driving adoption hardest, with usage rates well above the 55% average. But even older demographics are experimenting. The interface is simple enough — type a question, get an answer — that the barrier to entry is essentially zero.
So what are people actually asking about? According to HubSpot, the most common product recommendation categories include consumer electronics, skincare and beauty products, software and apps, kitchen appliances, and fitness equipment. Categories where consumers face overwhelming choice and high variability in quality. In other words, exactly the categories where traditional search has become most frustrating.
What This Means for Marketers — and Why Most Aren’t Ready
The implications for brands and marketers are profound and immediate. For two decades, the dominant playbook for product discovery has centered on search engine optimization. Get your product page ranking on the first page of Google. Bid on the right keywords. Build backlinks. That machinery isn’t obsolete yet, but its monopoly on consumer attention is cracking.
If a growing share of product discovery happens inside ChatGPT conversations — conversations that brands can’t see, can’t bid on, and can’t directly influence — the entire demand generation model needs rethinking. HubSpot’s research indicates that 65% of consumers who received a ChatGPT product recommendation went on to research the suggested product further, and 34% made a purchase based at least partly on the AI’s suggestion. That’s a conversion funnel operating entirely outside the traditional marketing stack.
The question every CMO should be asking: how does my brand show up in an AI recommendation?
The answer is complicated. ChatGPT’s product suggestions are shaped by its training data — which includes product reviews, editorial content, forum discussions, and publicly available web text. Brands with strong organic reputations, genuine positive reviews across multiple platforms, and consistent editorial coverage tend to surface more frequently. Brands that have relied primarily on paid media and SEO tactics without building authentic reputation signals are likely invisible to the model.
This creates a fascinating inversion. The brands that invested in genuine product quality, earned media, and community trust over the past decade are now being rewarded by AI systems they never anticipated. The brands that gamed Google’s algorithm with aggressive link-building and thin content are discovering those tactics carry no weight in a language model’s training corpus.
HubSpot recommends several tactical adjustments for marketers. First, prioritize building what they call “AI brand visibility” — ensuring your brand appears in the types of content that AI models train on, including authoritative reviews, expert roundups, and community forums like Reddit. Second, focus on product differentiation that can be articulated in natural language. ChatGPT doesn’t show images or video; it describes products in words. If your product’s advantages can’t be clearly stated in a sentence or two, the AI may overlook it in favor of competitors with more articulate value propositions.
Third, and perhaps most importantly, brands need to start monitoring how AI models represent their products. Several startups have emerged in recent months offering “AI search monitoring” tools that track how frequently and favorably a brand appears in ChatGPT and other AI assistant responses. It’s a nascent category, but one that’s growing fast.
OpenAI has also been making moves that accelerate this trend. The company has been expanding ChatGPT’s ability to browse the web in real time, access current pricing data, and link directly to retailers. The integration of shopping features — including product images, prices, and direct purchase links — signals that OpenAI sees commerce as a core use case, not a peripheral one. Recent reports from The Verge confirmed that OpenAI rolled out enhanced shopping features in ChatGPT, allowing users to see product cards with images, reviews, and pricing directly within the chat interface. No ads. At least not yet.
That “not yet” carries enormous weight. OpenAI is currently positioning ChatGPT’s shopping experience as ad-free and unbiased — a deliberate contrast with Google. But the company needs revenue. The temptation to introduce sponsored product placements will be immense. And when that happens, the trust advantage that currently drives consumer preference could evaporate quickly.
For now, though, the competitive dynamics are clear. Google has responded by accelerating its own AI integration through Search Generative Experience and Gemini-powered results. Amazon has introduced its own AI shopping assistant, Rufus. But consumer behavior data suggests ChatGPT has a meaningful head start in perceived trustworthiness, largely because it entered the space without commercial baggage.
There are legitimate concerns about the accuracy of ChatGPT’s product recommendations. The model can confidently suggest products that are discontinued, misstate specifications, or recommend items based on outdated pricing. HubSpot’s research acknowledges this: 28% of users who tried ChatGPT for product recommendations reported receiving at least one suggestion that was inaccurate or irrelevant. That’s a nontrivial error rate. But it’s apparently not high enough to deter repeat usage, particularly among younger consumers who’ve grown accustomed to treating all online information as provisional.
The broader pattern here extends beyond shopping. It reflects a fundamental change in how people interact with information. The ten-blue-links model of search, dominant for a quarter century, assumed that users wanted to choose from a curated list of sources. AI assistants assume the opposite — that users want a single synthesized answer. For product recommendations, the synthesized-answer model turns out to be what most consumers wanted all along. They never liked scrolling through pages of results. They just didn’t have a better option.
Now they do. And the data suggests they’re not going back.
Marketers, publishers, and retailers who built their businesses around traditional search traffic face a strategic inflection point. The consumer attention that once flowed predictably through Google is fragmenting across AI interfaces, social platforms like TikTok (where product discovery via short video is also surging), and conversational agents. The playbook that worked from 2005 to 2023 is losing relevance faster than most organizations have acknowledged.
The companies that will thrive in this environment share a common trait: they’ve built brands that people talk about in authentic, specific, positive terms — because that’s exactly the kind of language AI models absorb and reproduce. The rest will need to start building that reputation now, or risk becoming invisible in the channels where tomorrow’s purchase decisions are being made.
WebProNews is an iEntry Publication