Practical, data-backed guides on getting your store seen — and bought from — by AI assistants. Written from what our own audits and the AI-traffic index actually find.
What llms.txt is, why an online store needs one, and how to build in half an hour a file that helps ChatGPT and Perplexity understand your catalog.
llms.txt in WooCommerce two ways: the Botmetria plugin builds and updates the file from your catalog, or you place it by hand. Where it lives and how to check.
GPTBot, PerplexityBot, ClaudeBot: who to allow in robots.txt, why never to block Googlebot or Bingbot, and why the User-Agent alone is not enough.
Botmetria index: ~1,005,146 AI-bot requests hit 3 stores in two weeks. When it strains your server, how to see your own load, and how to throttle crawlers.
Find out whether GPTBot visits your store: grep your server logs, why a User-Agent alone is fakeable, and verify hits against OpenAI's IP ranges.
How to mark up schema.org/Product with offers price, priceCurrency and availability so ChatGPT and Perplexity can name your price, currency and stock.
ChatGPT says your product is sold out when it isn't? The culprit is the availability field in JSON-LD. Mark up stock with schema.org and keep it live.
Make shipping and returns readable for AI agents: dedicated pages, schema.org shippingDetails and merchantReturnPolicy markup, plus llms.txt links.
7 reasons ChatGPT, Perplexity and Google AI Mode can't see your store — robots.txt, 403s to bots, missing JSON-LD, SPA, no llms.txt — with checks and fixes.
Is ChatGPT quoting an old product price? We break down the causes — JSON-LD, feed and cache drift — and show how to keep prices fresh for AI assistants.
We audited 63 MD/RO stores (median 38/100). The top point-killer: product images returning 403/406 to AI bots. How to find the cause and fix it.
ChatGPT orders land in (direct) in Google Analytics because the referrer is stripped. Why GA undercounts the AI channel and how to actually attribute orders.
ChatGPT, Perplexity, Gemini, Copilot and Claude tag visits with referrer hosts — or not at all. How to label the AI channel and stop losing orders to 'direct'.
The AI agent's path from reading the catalog (UCP/ACP) to a checkout session where a human pays. What your store needs, the /.well-known/ucp manifest and a
MCP, ACP and UCP in plain words: how the agentic-commerce protocols differ, why your online store needs them, and how Botmetria supports each.
A practical 12-point AI SEO (GEO) checklist: robots for AI bots, llms.txt, JSON-LD price and availability, no 403 on images, and AI-channel attribution.
How to get recommended by ChatGPT and Perplexity: indexability, JSON-LD, llms.txt, authority and fresh data — a five-step system with real numbers.
A product feed for AI agents: XML, YML and agentic.json formats, the key fields id, price and availability, and how Botmetria turns your catalog into a UCP
Why AI agents need a clean markdown version of your product page, how to set up Accept: text/markdown and .md mirrors, and how to lift the markdown score.
How the 0-100 AI-readiness score is calculated: the five weighted categories — discovery, product, logistics, hygiene, Markdown — and how to read the bands.
