The 3 product categories that will win agentic shopping (and the ones that won't).

I started Flockd assuming AI shopping would lift any decent Shopify store. Two weeks of research and conversations with health, beauty, and supplement founders later: that's wrong.

Some categories are structurally built for AI shopping. Others will spend money optimizing for it and barely move.

Here's the data on which is which.

The framework

Three things need to line up for a category to benefit from AI shopping:

1. High consideration. Shoppers research before buying. They ask AI questions instead of typing brand names they already know.
2. Spec-driven. Products can be described in concrete attributes (dosage, ingredients, dimensions, certifications). AI extracts these and matches them to queries.
3. Fragmented market. No dominant brand. AI has to evaluate options, not just confirm what shoppers were going to buy anyway.

When all three are present, AI becomes the deciding factor in the purchase. When even one is missing, the whole channel breaks down.

The numbers worth holding onto

Before category specifics, here's the anchor:

• AI traffic average conversion: 2.47%
• Google Shopping: 1.95%
• Google Ads: 1.82%
• Average Shopify store (all traffic): 1.4%

AI shopping traffic already converts better than every paid Google channel. And that's just the average.

The top categories aren't slightly better. They're 2-3x the average, hitting numbers that rival email marketing.

The 3 categories that will win

1. Beauty and skincare

5.36% conversion from AI traffic. Highest of any vertical.

For context:

• 4x the average Shopify store conversion rate (1.4%)
• 3x the conversion rate of Google Ads (1.82%)
• Roughly the same as email marketing, the gold standard for paid channels

OpenAI named beauty as a category where ChatGPT Shopping Research "performs especially well." NielsenIQ found 49% of consumers already get beauty recommendations from generative AI.

Why it works: a query like "best vitamin C serum for sensitive skin under $40" maps directly to product attributes. Concentration percentages, ingredient lists, skin-type compatibility, certifications, pH levels. All of it extracts cleanly into the kind of factual specs AI agents use to rank products.

The opening for indie brands is real. EMARKETER's Q1 2026 AI Visibility Index showed brand mention rates declining across categories. AI is distributing recommendations across more brands, not consolidating around the giants.

2. Supplements and wellness

4.68% conversion from AI traffic. Second highest.

That's 2.5x the typical Shopify store and roughly double what supplement brands see from Google Ads.

230 million people ask ChatGPT health questions weekly (per OpenAI). When AI picks a winner, 74% of users follow the recommendation without checking alternatives.

Supplements are pure spec decisions: dosage, form, certifications (NSF, USDA Organic, third-party tested), ingredient origin, allergen info, dietary compatibility. A query like "best magnesium glycinate for sleep, third-party tested, no fillers" maps directly to data fields most supplement stores already have but don't structure properly.

The catch: AI applies YMYL (Your Money or Your Life) scrutiny to health products. They look for clinical data, third-party certifications, transparent sourcing, and editorial coverage from trusted health publications before citing a brand.

This makes supplements more of an opportunity, not less. Higher bar, more defensible position when you clear it. Market is fragmented, no clear leader, AI has to actually evaluate.

3. Sports and outdoor

OpenAI named sports and outdoor as a category where ChatGPT Shopping Research "performs especially well."

The category is built for AI matching. Waterproof rating in mm. Fabric weight in GSM. Pack volume in liters. Temperature rating in degrees. Fit type. Weight in grams.

Outdoor enthusiasts are notorious researchers. They read reviews, compare specs, watch comparison videos. AI accelerates a behavior they already do.

Market structure helps: big names dominate one tier (Patagonia, North Face, REI), but a deep middle layer of mid-sized technical brands competes for the buyers actively trying to escape the obvious options. If you make hiking poles, packable down jackets, ultralight backpacks, or technical running gear, this is your category to win.

Where AI shopping might be overrated

Fashion and apparel. 2.40% conversion from AI traffic. That still beats Google Ads (1.82%) and Google Shopping (1.95%), but it's roughly average for the channel. Don't expect outsized lift here. Aesthetics, fit, and social signaling drive these purchases, and AI can't help much with any of those. AR try-on tools will likely matter more than chat-based AI.

Single-product brands. People don't ask ChatGPT to recommend an iPhone. They just buy it. If your shoppers know your brand by name, AI doesn't enter the funnel.

Pure commodity products. Generic phone cases, basic t-shirts, plain candles. Compete on price, not specs. AI has nothing to differentiate.

Categories requiring physical try-on. Glasses, fitted shoes, makeup color matching. AI narrows options. The body still makes the decision.

Worth watching: pet products. Hits all three framework boxes but the data isn't there yet. Heavy emotional and recommendation-based decisions are harder to systematize. Revisit in 6 months.

What I'm watching next

A few things I don't have answers to yet, but am tracking as I keep building:

Whether the early winners stay winners. Right now beauty brands are converting at 5.36% from AI traffic partly because so few brands are optimizing for it yet. As more brands wake up to this, will smaller indie brands keep getting recommended, or will the big players with bigger budgets edge them out?

How fast OpenAI's direct merchant program changes things. OpenAI now lets merchants submit their products directly to ChatGPT instead of being scraped from the web. The brands signing up early are getting a head start on visibility. The longer other brands wait, the harder it gets to catch up.

What happens to fashion when AI gets better at images. Chat-based AI doesn't help fashion much because shopping clothes is visual. But Google Lens-style tools that recognize products from photos are improving fast. Fashion might bypass chat entirely and win in a different kind of AI shopping.

Whether pet products break into the top tier. They have all the right ingredients (people research, products have specs, no clear dominant brand). The data just isn't there yet. If it shows up in 6 months, it'll be the most underrated AI shopping category right now.

What's next

Next week I'm digging into how OpenAI, Perplexity, Google AI Mode, and Microsoft Copilot rank products differently. Brands that win on one platform sometimes lose on another. Worth understanding if you're optimizing across channels.

If you want to see how your store currently scores against AI shopping signals, that's what Flockd is for. Install it on Shopify — 14-day free trial, no credit card required.

See you next week.