How Revvye scans, scores, and surfaces revenue leaks.

When a scan runs, Revvye fetches a set of public pages from the submitted website, looks at the rendered output, and inspects the same supporting signals a non-logged-in visitor would meet: the homepage, top service pages, location pages where present, the public booking path, on-page schema, sitemap and robots, the public Open Graph and meta surface, the visible review and trust proof, and the public references that link back to the business from elsewhere on the open web. The scan is intentionally read-only. It does not submit forms, place test bookings, or attempt to reach gated pages.

The scan also runs a short visibility probe against the public behavior of the major AI crawlers and search crawlers. That is not a guarantee that a given answer engine is using your pages today. It is a check of whether your public posture would let that crawler in if it tried, plus a sanity check on the robots.txt directives, sitemap exposure, and rendering parity that decide whether a public page can be read at all. We treat that probe as one input, not a promise.

Everything Revvye uses is something you, a competitor, or a well-behaved bot could already reach. That is the intentional boundary of the product. It also means the scan can be run honestly without pretending to know anything about your private CRM, your booking platform, your payment data, or the inside of your business. That separation is part of how we keep the diagnostic legitimate.

Every issue Revvye finds carries a severity tier. Severity is what the math actually uses. The score is then expressed on a 0-to-100 scale where higher means fewer detected leaks across the public path and the visible trust surface, and where the individual category scores show you where the weight is sitting. We deliberately keep the category scores visible so that an operator can see, for example, that visibility is healthy but booking is leaking. That is the more useful read than a single top-line number.

The score is comparative against itself over time more than against other businesses. Two clinics in two different markets can both score in the same range and still be losing to completely different leaks. The point of the score is to make the next fix obvious for your specific public surface, not to rank you against a competitor you have not vetted. Where the report does compare to public competitors, that comparison is based only on the public surfaces of those competitors, with the same boundaries.

The major AI answer engines, including the public crawlers behind well-known generative search experiences, look at public pages a lot like a search engine does. They respect robots.txt directives, they hit the same rendered HTML, and they cite sources that they were able to read. If your robots.txt blocks those user agents, or if your site renders one thing for users and another thing for bots, your business is unlikely to show up cleanly in the answers those tools generate, no matter how good your service is.

Revvye checks the public posture against those crawlers as a proxy for “can an answer engine read your site at all?” If the answer is no, that is a hard ceiling on visibility in the answer layer of search. If the answer is yes, that is the start of the opportunity, not the end. Showing up in an answer also depends on entity clarity, schema, on-page evidence that your business actually does the thing it claims to do, and trust signals from outside the website. Revvye reports on those too.

We do not scrape any answer engine, and we do not claim to know whether a specific engine is going to cite you. What we report is whether your public pages are configured in a way that does not prevent that from happening. That is a real, useful, and honestly bounded signal.

The decision to keep the scan on public surfaces is deliberate. It means the scan is reproducible by anyone, including the competitor that decides to run Revvye against you. It means we never store credentials, never touch private records, and never create a security hole for the businesses we are trying to help. It also means the report is a fair description of what a real prospect sees when they evaluate you cold.

The price of that decision is real and we name it: Revvye does not see what your owner-operator dashboard sees. It does not know that you closed twenty-six bookings last week. It does not know that your phone is ringing, or that an email funnel is quietly working. The scan describes the public path. The operator owns the truth of what happens after that.

A free scan reflects the public surface at the time it ran. If you publish new pages, change your booking flow, ship schema, or update review prompts, those changes will not appear in an old scan. We recommend a follow-up scan after meaningful work ships, plus a recurring scan cadence when the operator wants to watch the public surface drift over time. Revvye Monitor exists for that recurring use case, but a manual rescan is always available.

The right cadence depends on how often the public surface changes. A clinic that updates providers, hours, and services every few weeks should rescan more often than a professional firm whose service pages barely move. The free scan is designed to be fast enough that the cost of running it again is low.

A leak is a place on the public path where ready-to-buy demand quietly exits. Maybe the homepage does not say what you do. Maybe the booking link is broken on mobile. Maybe the answer engine cannot read your service page. Maybe the reviews on the one page that gets traffic are five years old. Maybe the follow-up never happens because there is no email capture. The leak does not have to be dramatic. It just has to be costing you bookings you would otherwise have closed.

Revvye is built to point at those leaks one by one and to rank them by severity so the next fix is obvious. That is the entire product. The report is the long version. The scan is the short version. The methodology above is how we make sure the short version is honest.