Shopworthy

For online stores · 4 AI assistants

When shoppers ask ChatGPT what to buy, does it recommend you?

Check your store free. See what AI shopping assistants can read on your site, which brands they name instead of you, and the one fix to ship first.

More options
Competitors (up to 3: a website or a brand name)

Buyer questions are built from this category.

Live answers are on for ChatGPT, Claude, Gemini, Perplexity. Each report runs a small panel through their official APIs. Every key answered a test question.

  • Free
  • No signup
  • Results in about a minute
An illustration of the first card in every report: one buyer question in your category, asked again and again, with the brands named instead of you and the reason. Brands shown are invented.

Shoppers ask things like

  • “best soy candle for a gift”
  • “handmade letterpress wedding invitations”
  • “dog collar for a big puppy”
  • “vitamin C serum for sensitive skin”
  • “ultralight tent under 2 lb”
  • “light roast coffee subscription”
  • “linen sheets that stay cool”
  • “gluten-free granola with no added sugar”
  • “kids rain boots that are easy to put on”
  • “ceramic mug that keeps coffee hot”
  • “small-batch hot sauce gift set”
  • “running shorts with a phone pocket”

See a full report

Ridgeline Coffee is a fictional store we built to show every part of a report: the grade, the answers assistants give, the reasons and the fixes. Flip between before and after its fixes.

AI shelf certificateNo. 4VCH-TPLM-EK

Tested
2026-11-06
Method
rubric v2
Site
Ridgeline Coffee

Technical readiness · measured

Grade Fscore 31 / 100
Sample data, not measurements.

ChatGPT · sample answer

A shopper asks:“best light roast coffee subscription under $25”
  1. Summit Roasters
  2. Blue Fern Coffee
  3. Foxglove Coffee Works
  4. you · not on this shelf

Per assistant · sample answers

  • ChatGPT✕ Failnamed 0 of 12
  • Claude✕ Failnamed 0 of 12
  • Gemini! Needs worknamed 2 of 12
  • Perplexity✕ Failnamed 2 of 12

Basis: 48 sample answers: 4 assistants × 4 buyer questions × 3 runs.

Specimen certificate for Ridgeline Coffee, a fictional demo store. Its grade is measured; its answers are scripted samples.

Check, fix, prove

Each step shows its working. These fragments are the real interface, filled with the demo store.

  1. 01

    Check Free

    We put the questions your shoppers ask to AI assistants, more than once each, and count which brands get named. You see the answer itself, the tally, and the margin of error.

    How often each brand is named Sample
    • Summit Roasters33 of 48
    • Blue Fern Coffee22 of 48
    • Lowtide Coffee Co.15 of 48
    • Ridgeline Coffee4 of 48

    Bars show the share of 48 sample answers; whiskers show the 95% interval.

  2. 02

    Fix

    We fetch your store the way each assistant's agent does, find what turns it away or confuses it, and write the change: robots.txt, CDN rules, product markup, a product feed.

    Reason 1 · measuredChatGPT, Claude and Perplexity agents are turned away by your CDN or firewall

    Fix 01 /robots.txt

    # AI assistants and AI search: let them fetch and index pages for people (keeps you in their answers)
    User-agent: ChatGPT-User
    User-agent: OAI-SearchBot
    User-agent: Claude-User
    User-agent: Claude-SearchBot
    User-agent: PerplexityBot
    User-agent: Perplexity-User
    User-agent: Meta-ExternalFetcher
    User-agent: DuckAssistBot
    Allow: /
    
    # Search engines behind AI answers: Google AI Overviews and Gemini use Googlebot; Copilot uses Bing
  3. 03

    Prove

    Upload a server log and see which assistants really reached you, checked against their published IP ranges. On Pro, a timeline lines up each fix with assistant visits and AI-referred orders, and says “too little data” when that is the honest answer.

    Try the log checker

    AI-referred visits by week Sample · synthetic
    Fix shippedBeforeAfter

What AI shopping agents need from a store

Four things decide whether an assistant can recommend you and send the sale your way. We check each from the outside, the way an agent sees you.

The four pillars of AI shopping readiness
PillarThe questionWhat we check
1FoundCan AI shopping agents reach your store and your products?
  • Assistant agents allowed in robots.txt and at your CDN
  • Google and Bing can crawl and index your pages
  • A product feed for ChatGPT, Merchant Center for Google and Microsoft
2UnderstoodCan they tell what you sell, for how much, and whether it is in stock?
  • Product markup with price, currency and availability
  • Prices in your markup and feed match the page
  • Titles and attributes in the words shoppers use
3BuyableCan an agent take a shopper to checkout?
  • A checkout path that works without JavaScript
  • Return, shipping and privacy policies an agent can read
  • Agent checkout readiness: a UCP profile, an MCP catalog
4ProvenCan you see it working in your own numbers?
  • Verified assistant visits from your server logs
  • AI-referred visits from a small first-party snippet
  • AI-referred orders from an order export

Why now Reported by others, with the source under each figure.

Start free. Pay when you want proof.

  1. Free$0See what AI shopping assistants tell shoppers about your store, and why.
  2. Pro$79 / monthFor one brand that wants to get picked by AI shopping agents, and prove it paid off.
  3. Growth$299 / monthFor a brand with a team, several stores and a bigger competitor set.
  4. Agency$499 / month10 client stores included, then $39 each
  5. Fix pack$149 onceFor a small store that wants the fixes done once, without a subscription.

Compare plansFor agencies

Find out who the assistants name instead of you.