Skip to content
Free score

Can an agent get to your checkout?

Available now

Agent Conversion Score

Find out whether an AI agent can read, trust and buy from your site.

A 0 to 100 score of how an AI shopping agent experiences your site, split into Agent Choice (can it find, read, trust and recommend you) and Agent Checkout (can it buy), with your blocker and the fixes ranked by how much they move the score.

The free score runs live on this site in about a minute.

Get your free score

Pricing

  • Free score

    Free

    Live on screen in about a minute. Agent Choice, Agent Checkout, your blocker and top fixes by email.

  • Full report

    £249

    All 49 checks with evidence, the ranked fix list, and the same score for up to three competitors. £249, €279 or $299 depending on where you are.

  • Re-scan plan

    £79/month

    Monthly re-score, change alerts and a before-and-after on every fix.

Prices exclude VAT and sales tax where applicable. Monthly plans are billed monthly and can be cancelled any time. Refunds and cancellation.

Is this for you?

  • You sell online and cannot see how much of your traffic now comes through ChatGPT, Perplexity, Gemini or Claude.
  • Your site is behind Cloudflare or another bot-protection layer and nobody has checked what it does to agents.
  • You run on WooCommerce, BigCommerce, Magento or a custom platform, where agent readiness is not switched on for you.
  • You want a number to put in front of the board before you spend on an agentic commerce project.

Probably not for you if…

  • You sell only through marketplaces and have no site of your own.
  • You already publish a UCP profile, structured offers and a feed, and your checkout accepts agents. Run the free score to confirm, then skip ahead to the Simulated Agent Buyer.

What it measures

Access (25 points)
Whether GPTBot, ChatGPT-User, PerplexityBot and ClaudeBot can fetch your product pages at all, or get a 403, a rate limit or a "checking your browser" page.
Legibility (25 points)
Whether price, availability, identifiers, brand, delivery and returns are stated in structured data and in server-rendered HTML, so an agent reads them rather than guesses.
Offer clarity (15 points)
One canonical price that matches what the page shows, promotions expressed structurally, shipping cost and delivery time stated where an agent looks.
Transactability (25 points)
Guest checkout, a cart agents can reach, no captcha before payment, a UCP profile, an ACP document, an MCP endpoint and a product feed.
Trust (10 points)
Ratings with counts, policy pages linked from products, HTTPS without mixed content, a verifiable business identity and one consistent name.

How it works

  1. Paste your URL

    One field, no account. The audit starts immediately and runs live on screen.

  2. We fetch your site as five visitors

    A real browser plus the four agent identities, so access differences show up as data rather than guesswork.

  3. 49 checks run

    Robots rules, structured data, server rendering, policies, cart, checkout, well-known endpoints and feeds. Three checks cap the score if they fail, because they stop a sale outright.

  4. You get the score, then the breakdown

    Score and grade on screen for everyone. Add your email to unlock Agent Choice, Agent Checkout, your blocker and your highest-value fixes, on screen and by email. The full report carries the evidence for every check.

What you get

  • A score out of 100 and a grade from A to F
  • Two sub-scores, Agent Choice and Agent Checkout
  • Your blocker, if you have one, and the cap it puts on your score
  • A fix list ranked by how much each fix moves your score, with the exact change to make
  • Dimension scores for Access, Legibility, Offer clarity, Transactability and Trust
  • The same score for up to three named competitors (full report)
  • A before-and-after on every re-scan (re-scan plan)

What it does not do

Questions

Is this the same as an AI visibility tool?

No. Visibility tools count how often you are mentioned in AI answers. The score measures whether an agent that has found you can actually read your offer and complete a purchase. Most businesses need both; this is the half nobody else measures.

We are on Shopify. Do we need this?

Shopify publishes a UCP profile and agent checkout for you, which covers two of the 49 checks. The free score shows whether your theme, apps and bot settings undo that work. Many themes render price only with JavaScript, and many stores block agents through a security app without knowing.

What does the free score include?

The score and grade, live on screen in about a minute. With your email, also your Agent Choice and Agent Checkout scores, your blocker and your five highest-value fixes, on screen and by email. The full report adds the evidence behind every check and competitor scores.

How is the score calculated?

Two sub-scores. Agent Choice combines Access, Legibility, Offer clarity and Trust; Agent Checkout is Transactability. The headline works like a funnel, Choice × (0.7 + 0.3 × Checkout), so a site agents can recommend but not buy from is capped well below 100 without being zeroed, since people still buy through the handoff. Three checks cap the total when they fail. The full definition is published on the specification page.

Our pages show prices everywhere. Why does the score say there is no price?

Because the score measures what an AI agent receives, not what a person sees. Most agents fetch the HTML and read the structured data; they do not run your page's scripts. If the price is drawn in by script, it is on screen for people and absent for agents. The results page shows exactly what an agent received for each page we read.

Do you touch our site?

We fetch pages the way an agent would. We never add to a cart, submit a form, create an account or make a purchase.

Next in the suite

Simulated Agent Buyer — Where does an agent drop out of your funnel?

Get your free score.

Watch five visitors try your site, live. Your score in about a minute; the fixes by email.

Get your free score