What’s your website?
WUWI AI reads it the way a search engine does, then tells you what is holding it back. It takes about a minute, and you do not have to set anything up first.
How honest is the number
It reads 12 pages, and tells you what it could not see
WVGI is built from eight dimensions. A crawl can observe some of them almost completely and others barely at all, so the score is computed over the part that was actually measured — currently 40% of the index — and the confidence range widens to match. Nothing is modelled to fill the gap.
Search discoverability
25% visible · 18% of the score
Reads: whether pages are indexable and a sitemap is declared.
Answer & AI authority
45% visible · 14% of the score
Reads: whether AI crawlers may read the site, and whether the content is structured to be quoted.
Local reach
20% visible · 14% of the score
Reads: whether the site states a consistent address, phone number and opening hours.
Entity trust
45% visible · 14% of the score
Reads: whether the site declares who the business is in machine-readable form.
Content usefulness
60% visible · 13% of the score
Reads: titles, descriptions, headings, depth and structure of the pages crawled.
Technical accessibility
85% visible · 10% of the score
Reads: crawlability, indexability, response speed, mobile setup and internal structure.
Authority network
not visible to a crawl · 9% of the score
Reads: nothing yet.
Conversion readiness
50% visible · 8% of the score
Reads: whether a visitor has a way to call, email or submit an enquiry.
What it will not tell you
A crawl cannot see where you rank, how much traffic you get, or who links to you. Those need Search Console, an analytics account and a backlink provider. They are shown as absent rather than estimated, because a modelled number and a measured one look identical a month later — and only one of them is worth acting on.
No guarantee is made about rankings, traffic, citations, leads or revenue. Anywhere in the product.
What happens to your site
The crawler obeys your robots.txt and any crawl delay it declares, and reads up to 12 pages. It keeps metadata and a small ranked list of terms — never the text of your pages.
No result from this page is stored. There is no account to attach it to, and keeping anonymous crawl results against an address would be collecting data on people who have not asked for anything.
Every score is computed from the crawl by plain code — no model writes any of it. The figures above are read from the same module the product scores with, so this page cannot drift away from what it actually does.