AI Visibility for DTC
A shopper now asks an assistant "what's a good organic dog food under $40?" and gets three named brands. If yours isn't one of them, the problem usually isn't your product. It's that the machine can't read your store.
We ran an audit of 22 direct-to-consumer brands in 2026 to see how legible they actually are to AI assistants and answer engines. The headline number was not encouraging: the median AI-visibility score was 52 out of 100. Not a single brand in the set scored above 64. The "good" stores were mediocre, and the rest were close to invisible.
The most fixable failure was the most common one. 32% of the brands had zero product schema — no structured data telling a machine what a product is called, what it costs, or whether it's in stock. That's not a tuning problem. That's a store an answer engine has to guess about, and engines don't guess in your favor.
Two of those numbers deserve a second look. More than a quarter of brands (27%) were already losing to a named competitor on AI-discoverability — this is not a someday problem, it's a today one. And the median agent-commerce readiness was 86/100, meaning most of these stores could technically handle an AI-driven purchase. They just can't get found long enough to receive one. The plumbing is fine; the front door is locked.
The good news for a founder: the highest-leverage fixes are not a six-week project. Most of them ship in an afternoon, several in literal minutes. Here are the ones that move the score most.
This is the single biggest lever, because a third of stores are starting from zero. Product structured data (JSON-LD) is how you hand an assistant a clean fact sheet instead of making it scrape your HTML and hope.
At minimum, every product page needs a Product block with name, description, brand, image, sku, and an offers object carrying price, priceCurrency, and availability. If you collect reviews, add aggregateRating — assistants lean heavily on social proof when they rank options.
{
"@context": "https://schema.org",
"@type": "Product",
"name": "Organic Grain-Free Dog Food",
"brand": { "@type": "Brand", "name": "Your Store" },
"sku": "DOG-ORG-40",
"offers": {
"@type": "Offer",
"price": "39.00",
"priceCurrency": "USD",
"availability": "https://schema.org/InStock"
}
}Don't hand-write it and risk a typo that invalidates the block. Generate a valid one field-by-field with our free product schema generator, paste it into your product template, and confirm it parses with Google's Rich Results Test. Many Shopify themes ship partial schema — partial is not the same as valid, so check what's actually rendering rather than assuming the theme handled it.
A surprising share of "invisible" stores are invisible by accident: their robots.txt blocks the very crawlers that feed AI assistants, often inherited from an old SEO plugin or a copy-pasted config. If the bot can't fetch the page, nothing else on this list matters.
Open yourstore.com/robots.txt and read it. Make sure you are not disallowing the user-agents used by major AI and answer-engine crawlers. You don't need to name every bot — just make sure your product and collection paths are reachable and you're not sitting behind a blanket Disallow: / for a class of agents. This one's free to fix and instantly raises your ceiling on every other change.
An llms.txt file at your domain root is a short, curated map you hand directly to assistants: here are my key collections, here's what I sell, here's the canonical page for each. Instead of forcing a model to reverse-engineer your mega-menu, you give it a clean summary it can quote with confidence.
It's plain text, it takes minutes, and a large and growing share of stores still don't have one — which means it's cheap differentiation right now. Spin one up from your existing URLs with the free llms.txt generator and drop it at yourstore.com/llms.txt.
Assistants read words, not pixels. If your sizing chart, ingredient list, material, dimensions, or warranty live only inside a graphic, those facts effectively don't exist to a model — and those are exactly the details a shopper's question hinges on ("is it machine washable?", "does it fit a 15-inch laptop?").
Walk your best-selling product page and ask: if I deleted every image, would the description still answer the obvious questions? If not, move the spec facts into real, crawlable copy. Write the description for a literal-minded reader: concrete attributes, plain claims, no marketing fog. Clear factual copy is what gets quoted in an answer.
You can't fix a score you've never seen. Given that 27% of audited brands were already behind a direct competitor, the practical move is to measure where you stand before a rival pulls further ahead. The first four fixes above will move the needle for almost everyone — but you want to confirm it, not assume it.
Run your domain through our free AI Visibility Check to get a score and the specific gaps holding it down. Ship the fixes, then re-run it. A store starting from a median 52 can usually clear the schema-and-crawler basics and watch the number climb the same week.
Here's the part most founders miss. Median agent-commerce readiness was 86/100 across the audit — far higher than the 52 visibility median. Translation: most of these stores are technically capable of completing a sale initiated by an AI agent. What they lack is the discoverability to ever be chosen, and the machine doorway to be transacted through — recall that 91% had no MCP endpoint at all.
That gap is the whole opportunity. The expensive, complicated layer — being able to actually fulfil agent-driven commerce — is mostly already there. The cheap layer that decides whether you're in the consideration set is the one nobody's doing. Closing it is an afternoon, not a quarter.
For the full methodology and the brand-by-brand breakdown, read the State of AI Visibility in DTC report.
Two free tools and a 5-minute audit, in the order we'd run them ourselves.
Want it done with you? The AEO Audit + Fix List ($99) grades your store against the audit benchmark and hands you a prioritized, copy-paste fix list — so you ship the five moves above without guessing which one matters most for your catalog.
It's how easily AI assistants and answer engines can find, read, and correctly cite your products when a shopper asks for a recommendation. It depends on machine-readable structure: product schema, an llms.txt, a crawler-friendly robots.txt, and product facts written as real text. In our 2026 audit of 22 DTC brands the median score was 52/100, with no brand above 64.
Usually because the machine can't parse your catalog. The common causes are missing Product structured data, AI crawlers blocked in robots.txt, key facts trapped in images, and no llms.txt. In our audit, 32% of brands had zero product schema — which makes a store effectively unreadable to answer engines.
Add a Product JSON-LD block to your product template covering name, description, brand, image, sku, and offers (price, currency, availability), plus aggregateRating if you have reviews. Generate a valid block with the free product schema generator, paste it into your theme, and confirm it with a rich-results test.
It's a plain-text file at your domain root that gives assistants a concise, curated map of your most important pages. It takes minutes, and a large and growing share of stores still don't have one. The free llms.txt generator builds one from your existing URLs.
They overlap but aren't identical. SEO optimizes for ranking links; AI visibility (answer engine optimization) optimizes for being read and cited inside a generated answer. Structured data, factual clarity, and crawler access carry more weight here than keyword density or backlink volume.
Figures cited are from Hatchloop's 2026 audit of 22 DTC brands: median AI-visibility 52/100 (no brand above 64), 32% with zero product schema, 27% trailing a direct competitor on AI-discoverability, 91% with no MCP endpoint, and median agent-commerce readiness of 86/100. Full report: /state-of-ai-visibility-dtc/.