Free, public information only
Check what your storefront tells the outside world.
Enter a domain. We read the page your server sends, its response headers, one product page and your robots file, then report what is actually there: pixel, tag manager, consent platform, server-side endpoint, currencies, hreflang, payment markup, script weight and Core Web Vitals.
Where a question cannot be answered from out here, we say not verified and name what would answer it. We never turn a thing we could not see into a thing that is broken.
Three requests to your server, robots file honoured, roughly twenty seconds. Nothing is stored against you and nothing is emailed to you unless you ask.
| Area | Weight | Answered | Score | out of 100 |
|---|
Weights are ours and they are written down: see engine/score.mjs. Measurement carries 40 because everything downstream is decided by it. Ads efficiency carries 0 because nothing in it is visible from outside a store.
| Status | Check | Reading |
|---|
What this page cannot reach
Questions that have no public answer.
These are not gaps in the tool. They are gaps in what a storefront exposes to a visitor. Each one needs read-only access to an account you own, and each one is in the paid audit.
Method
What this actually does, in order.
1. robots.txt
Read first, obeyed. If your robots file tells us to stay out, we stop there and report that we stopped.
2. Your homepage
The HTML your server sends, before any JavaScript runs, plus the response headers and cookie names. That is where the tag layer usually lives.
3. One product page
The first product link on the homepage. Express checkout buttons and payment markup sit there, not on the homepage.
4. Core Web Vitals
Field data from the Chrome UX Report through the PageSpeed Insights API: real users over 28 days, not a lab score. If the API is unavailable, this reports as not verified.
Three requests to your server, one to Google. We do not add items to a cart, we do not submit a form, we do not create an order, and we do not run anything at checkout.
The four states
Why not verified is the useful one.
- passWe looked, we found it, it is what it should be.
- warnWe looked, we found it, and it is doing less than it could. Usually a fixed-scope job rather than a rebuild.
- failWe looked, we found it, and it is wrong. Every fail on this page names the evidence it rests on.
- not verifiedWe could not decide this from public information, so we are not going to pretend. This is the honest half of the report, and on most stores it is the larger half.
A tool that turns every question it cannot answer into a red alert produces an impressive looking page and a useless one. Most of what decides whether your reported ROAS is real happens inside Events Manager and inside Shopify, where no visitor can see it. Saying so is the point.
Next
The Growth Infrastructure Audit.
This page checks what is visible from outside your store. The audit checks what is actually happening inside your accounts: whether the Conversions API is deduplicating, what your Event Match Quality is per market, how far apart your Meta purchase count and your Shopify order count are over thirty days, which countries already have sessions and almost no orders, and how many of your checkouts that start actually pay.
This page
Public information. Three requests. Twenty seconds. Free. Answers whatever public markup can answer, and names every check it could not.
The audit
Read-only access to Meta, Shopify and GA4, plus a 12 question form. Six areas, 33 checks, every number traced to a query or a screenshot. Five working days, then a 45 minute call.
We are wrong for you below roughly 30,000 EUR a month in ad spend: the fix costs more than the misallocation it recovers, and we will say that on the call rather than take the project. The audit still stands on its own, because the cross-border and checkout findings do not depend on spend level.