US Shoppers No Longer Start on Google: How Sellers Get Found on the AI Shelf

What drives a US consumer's buying decision is no longer just a search ranking. It's a longer path: a product gets discovered on social media or in an AI assistant, checked on Google, then bought on Amazon or a brand's own site. For Chinese sellers, the job is to learn what US buyers in your category look at and trust at each step, then make sure they can find you and understand you at every one.

New data makes that path clearer. This month Search Engine Land covered a Stella Rising survey of 511 beauty shoppers: only 43% start looking for products on Google. The rest are spread across Instagram (11.9%), YouTube (11.2%), TikTok (10.6%) and AI tools (9.8%). But after discovering a product on TikTok, 72% go on to search Google, and only 7% buy directly on TikTok.

Meanwhile, Modern Retail explored the "AI shelf": more and more shoppers simply ask ChatGPT, Claude or Gemini to recommend products. Tech writer Deana Burke even invented a natural deodorant brand to test whether AI would recommend it, and it worked.

Where do US consumers start looking for products now?

Everywhere, but rarely in just one place. The survey respondents averaged 47 years old and all came from the beauty category, so be careful applying the numbers to other categories. Still, a few signals matter to every seller exporting to the US:

• Younger shoppers turn to AI sooner. Among respondents in their 20s, 24% start with Google and another 24% with AI tools, while 20% start on TikTok. Among those in their 60s, half start with Google.
• AI is now a routine step. 59.5% of respondents use AI at some point while searching.
• Google is still where people verify. When they can't find what they need on TikTok or YouTube, 61% go back to Google.
• Searches look more like questions. On TikTok, 24% search a product name, 19% a skin or hair concern, 18% a brand name, and another 18% type a full question.
• Most purchases happen on Amazon. 47.7% of respondents usually complete beauty purchases on Amazon.

What factors influence US consumers' buying decisions?

Put the survey next to the AI shelf idea, and at least four factors shape what US buyers do:

• Where they discover you. Social media and AI assistants are taking over the "first look." 41% of respondents bought a beauty product in the past six months because of TikTok or YouTube; among people in their 30s, it was 60%.
• Whether they can verify you. Once interested, buyers search your brand name on Google and read reviews. If they find nothing, or a mess, the earlier exposure is wasted.
• Whose word they trust. In the survey, 14.1% of respondents rated YouTube creators as highly trustworthy, versus 8.6% for TikTok creators. Trust isn't the same on every platform.
• How AI describes you. When a buyer asks AI directly, "Which one is right for me?", the list and the reasons it gives become a new kind of shelf space.

What is the "AI shelf," and why does AI recommend a brand?

Modern Retail frames it this way: product discovery moved from physical shelves to digital ones, and now there's an extra layer, the AI shelf. Instead of comparing search results one by one, shoppers ask AI to hand them a few picks.

Exactly how AI decides whom to recommend is something the discussion itself lists as an open question. But Burke's experiment shows at least that AI recommendations are shaped by the information available online; even a made-up brand got recommended. Flip that around: a real Chinese brand with a good product but sparse, inconsistent information online could easily be ignored by AI.

The work of getting your brand cited accurately in AI answers is called GEO, or Generative Engine Optimization. It starts with the truth: clear product specs, consistent brand information and verifiable reviews, never anything made up.

Why do sellers across the Pacific keep misreading US buyers?

Economists call it information asymmetry: one side of a deal knows more than the other. In cross-border selling, the gap runs both ways.

On one side, Chinese sellers know their products best but don't know how US buyers decide: which platform sparks their interest, what words they search, which creators they trust. On the other, US buyers know little about an unfamiliar Chinese brand, so they lean on signals such as reviews, creators and AI recommendations to judge whether it's worth buying.

ToucHigh's competitive-content research also found that, in the category it studied, more than 80% of AI answers couldn't be verified. If a brand doesn't supply clear, verifiable information itself, AI and everyone else will fill in the blanks for it. That's why understanding comes first: close your own information gap about buyers, and you can close their information gap about you.

How can Chinese sellers map the US buyer journey in one week?

Start with this table to see what buyers do at each stage and what you need to have ready:

1. List 20 questions real buyers ask. Mine Amazon's negative reviews and Q&A sections and TikTok's search suggestions, and write every question down in English.
2. Ask AI each one. Put them to ChatGPT, Perplexity, Gemini and Google AI Overviews, and record which brands get recommended and which sites get cited.
3. Search the same questions on social. Note which creators and formats hold the top spots and what people ask in the comments.
4. Google your own brand name. Does page one show your website and reviews, or unrelated results and even negative content?
5. Find the gaps against the table. Mark the stages where you're missing, and fix "verify" and "compare" first, since they affect sales the most.
6. Fill the gaps, then recheck. Write an FAQ page and comparison pages, make product information consistent across channels, and rerun the same questions a month later.

How should you split budget between social and GEO?

The Search Engine Land piece gives sound advice: don't move your whole social budget into GEO because you think AI is replacing search. The two work together. A strong social presence and outside word of mouth are themselves sources AI draws on when it recommends brands.

A more practical approach is to make visibility your main metric: do you show up in TikTok search results, how does AI describe your brand, and what share of organic conversation do you hold in your category? Accept the measurement blind spots, too. The article notes that TikTok's in-app browser may not fire GA4 (Google Analytics 4) events correctly, and Amazon sales don't show up in your independent DTC site's analytics, so don't judge channels on last click alone. If you want to know where your brand really stands in US AI answers, start by looking at ToucHigh's GEO Autopilot.