Does your site exist for AI agents?
◆ Free scan
What it measures
Six categories. Schema validators look at one page; Merchant Center looks only at the feed. What is different here is that access, schema, feed and freshness are assessed in one pass, as a chain: a site with a flawless feed that blocks crawlers cannot score well.
OAI-SearchBot and Google-Extended in robots.txt, the HTTP status agent user-agents receive, whether content renders without JavaScript, sitemap coverage.
Organization, Product and Offer schema, ISO 4217 currency, a valid availability enum, GTIN and MPN coverage, variant modelling.
Required fields, id uniqueness, GTIN coverage, price format, the availability enum, whether product links resolve.
Feed price compared against the live page across a sample, plus feed reachability.
llms.txt, target-market language and hreflang, factual density in the copy, FAQ and comparison signals.
GA4 and GTM detection, pixel and conversion measurement signals.
How it works
Scan
Enter the domain. Add your product feed URL if you have one.
Score
A 0-100 score across six categories plus a maturity tier, with the gate rules applied.
Fix
Your first five fixes ranked by weight, each one saying what to change.
Re-scan
Run it again after fixing. The score is deterministic, so a change is a real change.
What the score means
A tier is not decided by points alone. If the agent access category scores under 10 the total is capped at 40, and if feed and site prices disagree, 15 points come straight off.
| Score | Tier | What it means |
|---|---|---|
| 0-25 | Invisible | The agent cannot read the site or cannot understand the product. |
| 26-50 | Accessible | The agent can read it, but the data is not trustworthy. |
| 51-75 | Feed-Ready | The data is trustworthy, but there is no visibility or measurement. |
| 76-90 | Agent-Native | The channel is live and measured. |
| 91+ | Reference | Mature, governance and trust layers included. |
Who it is for
Exporter
The technical answer to "I sell into the US, so why don't I show up in ChatGPT?"
Agencies and freelancers
A fast, evidenced starting point for a client audit or a pitch.
In-house teams
A tier assignment and a prioritised first 90 days.
◆ Who it is not for
If you sell only domestically, have no export ambition and carry no product catalogue, part of what this tool measures currently returns nothing. ChatGPT Ads does not target Turkiye, and the first wave of agentic checkout has been pulled back. The scan still helps on access, schema and measurement, but think twice before investing in the feed and protocol items.
Frequently asked questions
◆ What does the score not guarantee?
It does not guarantee sales. The score measures technical readiness, not commercial success. A high score means agents can read your site and understand your product correctly; it does not mean they will choose you.
◆ Does the whole 148-item list feed the score?
No. Only the externally verifiable items produce the score. Contracts, approval thresholds and team training cannot be scanned; they stay on the manual checklist so they cannot inflate the number.
◆ I sell services, not products. Is this useful to me?
Yes. If no product pattern is found in the sitemap the site is scored as a service or B2B business: the feed and freshness categories drop out of the denominator, and the structured-data checks look at Service, FAQPage and BreadcrumbList instead.
◆ Will the same site score the same twice?
Yes, it should. No language model computes the score; the checks are deterministic. That is what makes a before-and-after comparison meaningful.
◆ Will the scan load my site?
No. It fetches the homepage, robots.txt, llms.txt, the sitemap and at most six pages sampled from it. Our crawler identifies itself as `WebturesAgentReadinessBot` and links back to this page.
◆ What if my site blocks the scan?
That is itself a finding and the report says so. If your WAF blocks a transparent, robots-respecting audit crawler, it is very probably blocking agents too.
◆ Honest limits
- Agentic commerce today is mostly a discovery and visibility channel. The first wave of embedded checkout was pulled back for behavioural reasons; most of the investment belongs in feed quality and AI visibility.
- This scan measures what is visible from outside. Stock accuracy, post-commission margin and approval thresholds can only be verified internally.
- ACP, UCP and AP2 specifications change quarterly. The checks will hold; the technical detail will not.
- Claims like "AI traffic converts X times better" are usually single-source and selectively sampled. Your own baseline is the only evidence that counts.