• AI visibility

Guide · AI search

How can an apparel brand win new customers through AI search?

By making the facts that decide a clothing purchase, fit by size, fabric, care and returns, complete and consistent everywhere AI tools look, and by earning independent mentions that back them up. AI is not yet where most Americans start looking for clothes, but it is where many check fit, compare products and narrow a choice. For an apparel brand, the payoff is a new customer who keeps what they bought and comes back.

By Underneath Team Guide Updated

The short version

  1. Fit is the costly problem: Coresight Research estimates (opens in a new tab) the US online apparel return rate reached 23.4% in 2025, in an online apparel and footwear market worth $201.1 billion, and nearly 70% of shoppers who returned clothes cited size and fit.
  2. Discovery through AI is still small in the US, help with fit is not: in a YouGov poll (opens in a new tab) of 957 US clothes shoppers, only 6% wanted to use tools like ChatGPT or Gemini to discover clothing, but 25% wanted AI size and fit recommendations.
  3. Product facts do not line up across channels: only 14% of UK shoppers in the Connected Consumer 2026 report (opens in a new tab) said fashion product details always match across social platforms, marketplaces and retailer sites.
  4. AI answers favor known names: EMARKETER’s index (opens in a new tab) of ChatGPT recommendations across 20 apparel and fashion categories found Nike mentioned in 12% of them in August 2026.
  5. Specific brands can still win: Shopify reports (opens in a new tab) that Cottonique, a hypoallergenic apparel brand, grew AI-referred sales 276% year over year.

Who buys everyday apparel, and what is a new customer worth?

Shoppers replacing basics they wear constantly; a customer who finds a good fit tends to buy the same item again.

Everyday apparel, the T-shirts, jeans, underwear, hoodies and work pants people wear every week, is bought differently from trend-led fashion. The shopper usually knows what they need. The questions are whether it will fit, how the fabric feels and lasts, how to care for it, what it costs and whether returning it is easy. Trend-led style discovery is a separate question; this article is about the product facts.

The value of a new customer is in repeat purchases of something that fits. We found no public lifetime value figure for apparel brands, so we do not invent one. What is public is the cost of getting fit wrong. Coresight’s 23.4% return rate means that, for online clothing, roughly one item in four comes back, and size and fit cause most of those returns. A brand that wins a customer with an accurate answer, and keeps the sale, earns twice: the first order and the next ones. Whether AI can also lower what you pay for that customer is the subject of our guide to DTC acquisition costs.

Bodies and needs are also changing. In a Coresight survey reported alongside its sizing study, 70% of US GLP-1 users said they had dropped at least one clothing size. Every size change is a moment when a shopper looks for something new and asks what will fit. Shoes raise the same fit questions; see how footwear brands get found.

Where does AI already sit in how people shop for clothes?

Mostly in research and fit questions, not yet as the main place Americans discover clothing.

The evidence is mixed, and the mix is useful. YouGov’s May 2026 poll found US clothes shoppers prefer to discover new styles and brands by browsing in stores (60%), on retailer websites or apps (46%), through friends or family (40%) and through search engines (37%). Interest in general AI tools for discovery was just 6%. But a quarter wanted AI help with size and fit, and 26% wanted it to check stock. Their biggest worries were privacy (51%) and the accuracy of recommendations (47%).

Other surveys find more use. In the UK, the Connected Consumer 2026 report, based on 2,000 consumers, found 60% use tools such as ChatGPT, Claude or Gemini at least occasionally while shopping for fashion, mostly to find deals, compare products and summarize reviews. Coresight found 58% of US consumers familiar with AI had used or intended to use AI tools for shopping. In Global Payments’ survey of more than 16,000 consumers, reported by FashionUnited (opens in a new tab), 42% worried an AI agent could buy the wrong item, which the report notes is a particular risk for clothing because fit and sizing vary between brands.

The AI shopping tools themselves are built around clothing. Google says its AI Mode shopping experience (opens in a new tab) draws on a Shopping Graph of more than 50 billion product listings, with more than 2 billion refreshed every hour, and lets shoppers virtually try on billions of apparel listings using their own photo. OpenAI documents a “Try on” button for clothes and accessories in ChatGPT. For accessories such as jewelry, shoppers ask about certification, sourcing and gifts instead, as covered in how jewelry brands get recommended. Across all US retail, Adobe’s data (opens in a new tab) shows AI traffic rose 393% in the first quarter of 2026 compared with a year earlier.

Our reading: AI is a growing research and decision tool for apparel, used more for “will this fit and is it worth it” than for “show me something new.” That is exactly where product data decides the outcome.

Which questions do apparel shoppers ask AI?

Questions about fit, fabric, care, specific problems and returns, often in the shopper’s own words.

We wrote the prompts below to illustrate the questions; they are not observed queries:

  • Fit: “Jeans for athletic thighs and a small waist, 32-inch inseam, that don’t gap at the back.”
  • Fabric: “Heavyweight 100% cotton T-shirt that won’t shrink or twist after washing.”
  • Care: “Work pants I can machine wash and wear without ironing.”
  • Problem: “Wire-free bra for sensitive skin that doesn’t itch.”
  • Size change: “I dropped two sizes; which brands run true to size in women’s trousers?”
  • Returns: “Which basics brands offer free returns and exchanges?”

The problem-style question matters most for smaller brands. Shopify quotes Cottonique’s co-founder: “Nobody with eczema types ‘hypoallergenic bra.’” Its report describes how an AI that can parse “my bra makes my skin burn” needs a brand that actually solves that problem, and Cottonique’s AI-referred sales grew alongside a 52% rise in total sales. That is one brand’s reported experience, not a general result, but it shows how AI answers can connect a plain-language need to a specific product.

How does an AI answer turn into an apparel sale that sticks?

Through a named product, a listing with the right size and price, a try-on, a purchase, and no return.

The answer. The assistant or AI Mode panel names a few products. Google says each listing in its Shopping Graph carries details like reviews, prices, color options and availability.

The listing and try-on. The shopper sees sizes, prices and stock, and may try the item on virtually. OpenAI warns that try-on images “do not guarantee fit or size” and tells shoppers to check the merchant’s measurements, product details and return policy. Those measurements are your data.

The purchase. It may happen on your site, at a retailer or marketplace, or through agentic checkout: Google describes tracking a price and tapping “buy for me” to complete checkout on the merchant’s site with Google Pay.

The keep or return. This is where apparel differs from most categories. With size and fit behind most clothing returns, a reasonable expectation is that answers based on accurate measurements lead to fewer returns, and answers based on vague size labels lead to more. We found no published data on return rates for AI-referred apparel orders.

The repeat. A basic that fits is bought again. Aviator Nation’s ecommerce director told Shopify that someone might first meet the brand in person, then use its site, social media “or ChatGPT to find the product afterward.” AI is one touchpoint in a longer relationship, not the whole of it.

What decides which clothing brands AI names?

Platforms document product data, reviews and price; studies show known brands and outside sources carry weight.

Documented by the platforms. OpenAI says ChatGPT considers structured metadata from first-party and third-party providers, such as price and product description, along with other third-party content and reviews, when choosing products. Google’s Merchant Center rules (opens in a new tab) make the size attribute required for free listings of clothing products, and Google recommends the material attribute (opens in a new tab) “if customers might search for your product by material or if they might decide to buy your product based on the material.” Google also says AI Mode uses query fan-out, running several searches at once to work out what makes a product right for a request.

Observed in studies. EMARKETER’s apparel index says AI recommendations “favor established brands,” that visibility varies sharply by category, and that retailer and publisher sources shape which brands surface. Our own Reddit citations study found Google’s AI cited Reddit in 17.9% of AI Overviews; community discussion of fit and durability is part of what these systems read. A separate piece asks whether Reddit shapes Google AI Overviews.

Industry argument, not platform documentation. Coresight’s report argues that brands with inconsistent or incomplete sizing data “risk becoming less visible” as AI agents recommend products. That is a research firm’s view, written with a sizing technology company, not a documented ranking rule. We think it is a reasonable expectation, because assistants can only match a fit question to fit data that exists. Makeup brands face the same matching problem with shade, as makeup recommendations from AI shows.

Trust factors specific to apparel. Accurate size charts with garment measurements, honest fit notes (runs small, relaxed through the hip), fabric composition and weight, stretch, shrinkage and care instructions, clear return and exchange terms, reviews that mention fit and how pieces hold up after washing, and editorial and community mentions that confirm all of it. Our article on what drives AI product recommendations covers the wider evidence.

What does an apparel brand lose when AI leaves it out or gets it wrong?

New customers at the moment of a fit question, and margin when wrong answers lead to returns.

If an answer names three brands for “jeans that don’t gap at the waist” and yours is not one of them, you lose a shopper who might have stayed for years. We infer this from how answers narrow choices; no public data measures it for apparel. If an answer names your product with wrong size or fabric facts, the cost arrives later, as a return. The inconsistency problem is real: with only 14% of UK shoppers seeing product details that always match across channels, a brand’s marketplace listings, retailer pages and own site often disagree, and an assistant may repeat whichever version it finds. AI-driven sales can also be hard to see; our article on why analytics miss AI visibility explains why.

How does GEO work for an apparel brand?

It makes your fit, fabric and returns facts easy for AI to find and trust, on your site and elsewhere.

For a clothing label, generative engine optimization (GEO) means getting AI answers to name your pieces and describe their fit, fabric and care correctly. No one can buy or promise a recommendation. For apparel it usually covers:

  1. Fit data by size. Garment measurements for every size, fit notes and model sizing, on product pages and in feeds, not only in a generic chart.
  2. Fabric and care facts. Composition, weight, stretch, shrinkage and care, written the same way everywhere. Our guide to what product content AI shopping assistants prefer explains why specific facts beat adjectives.
  3. Complete feeds. Size, material, color and availability attributes filled in for Google and other channels, kept current.
  4. Consistency across channels. Matching facts on your site, retailers and marketplaces, so assistants do not see conflicting versions. We weigh the marketplace side in whether AI search reduces dependence on marketplaces.
  5. Answers to problem questions. Plain pages that use the words customers use, such as itchy seams, gaping waistbands or shrinking hems, and say honestly which products solve them.
  6. Reviews and community. Honest customer reviews that mention fit and wear, and genuine participation where people discuss basics. Never fake reviews or planted posts.
  7. Independent coverage. Editorial roundups and product tests earned through digital PR, so the facts are confirmed by someone other than you. Our article on which pages to target explains why.
  8. Measurement. Tracking a fixed set of fit, fabric and problem questions across ChatGPT, Google AI Mode, Gemini and Perplexity over time.

What don’t we know yet about AI and apparel sales?

How much apparel revenue AI drives today, and whether AI-referred orders are returned less often.

The surveys disagree on adoption: 6% interest in AI discovery among US clothes shoppers in YouGov’s poll, against 60% occasional use for fashion among UK consumers in a report from a company that sells search and merchandising software. The Coresight sizing report was produced with a sizing technology partner. EMARKETER’s index covers ChatGPT only. Cottonique’s and Aviator Nation’s experiences come from Shopify, which sells the tools behind them. Adobe’s traffic figures cover all retail. We found no independent data on conversion or return rates for AI-referred apparel orders, and no platform documentation that says size data affects ranking directly.

Where should an apparel brand start?

With your best-selling basics: check what AI tools say about their fit, fabric and returns, and where they are wrong.

Pick 20 to 30 questions your customers ask about fit, fabric, care and specific problems, and run them across the main AI assistants and Google’s AI Mode. Record which brands are named, how your products are described, which sizes and measurements are quoted and which sources are cited. Then compare your own site, feeds and retailer listings for conflicts. If you would rather hand that off, ask us to audit how AI describes your range; we will set out the product-data, review and coverage work most likely to bring in new customers who keep what they buy. Our generative engine optimization service page describes how that work on fit data, product feeds and channel consistency is planned and then measured.

Frequently asked questions

Do AI shopping tools let customers try on our clothes?

Yes, on some surfaces. Google offers virtual try-on across billions of apparel listings in Search, and ChatGPT offers a “Try on” button for clothes and accessories. Both work from product images, so good images and accurate measurements matter.

Should we focus on our own site or on marketplaces?

Both, with the same facts. Shoppers and assistants see your products in many places, and inconsistent details are common.

Is this different from fashion trend discovery?

Yes. Trend discovery is about style and inspiration. Everyday apparel turns on fit, fabric, care and returns, which are matters of accurate product data.

Can a smaller apparel brand compete with Nike in AI answers?

On broad questions it is hard. On specific needs, such as sensitive skin or unusual fit, a brand that clearly solves the problem has a better chance. Our look at whether AI assistants favor big brands covers the wider pattern.

Sources

Free strategy call

Some questions are easier to answer about your own business.

Bring the one that matters most. On a free 30-minute call we’ll take a first look at it and send you a short written read afterward.

Ask about my businessHow we work