We measure first and talk after.

The scan works like a browser, not like a checklist. It requests your page, follows every redirect, and reads what actually came back. Each number in the report comes from that one response. Nothing is estimated and nothing is measured in a laboratory. If your server was slow at that moment, the report says so, and it says how slow. The same scan runs on any public address, so you can point it at a competitor just as easily as at your own site.

You get two numbers because one is not enough. The first is what the page scores today. The second is what it scores once the task list has been worked through. That second number is deliberately not a hundred. A score that is almost always perfect measures nothing, so we assume the work gets done well rather than flawlessly.

The full audit goes further than a single page. It reads up to thirty of them and finds what one page cannot show: titles that repeat, internal links that lead nowhere, two addresses serving the same content. Every finding names the page, the code and the change. Each one is marked as work for a developer, for an editor, or for whoever manages the hosting.

What the scan measures

Eight categories, each weighted by how much it moves the needle: content quality, technical and indexing, on-page SEO, structured data, speed and mobile, AI findability, security and images. The norms come from what search engines document publicly - their own guidance on titles, structured data and page experience - not from a house opinion. Where a finding rests on a published rule, the report says which one.

What it costs

The basic scan is free and asks for nothing. A full audit of up to thirty pages is a fixed amount, well under what agencies charge on average. The pricing page lists the four options side by side with the market figures next to them.

See what an audit costs

Why the second number is never a hundred

The score within reach assumes the task list gets worked through well, not flawlessly. A binary finding - a missing canonical, an absent header, no sitemap - really does get solved. A ratio or a band, such as alt texts or load time, lands on good rather than perfect. Depth of content, internal link structure and server speed keep asking for work and only shift part of the way. That is why a page that starts at forty does not land at a hundred: the number would stop meaning anything if it always did.

What this scan does not do

It reads the page as it was served at that moment, so it says nothing about your rankings, your backlinks or your competitors. It does not run your JavaScript, which means a page that builds its content in the browser looks emptier here than it does to a visitor - and, notably, emptier to a crawler that skips JavaScript too. It measures one page, and only the page at the address you gave. Those limits are deliberate: a number that covers everything usually measures nothing.

Which standards this follows

Structured data is validated against the vocabulary published at schema.org, robots.txt is read the way RFC 9309 describes it, and the machine-readable summary of this site follows the llms.txt convention. All three are open specifications, so you can check any finding against the source rather than taking our word for it.

schema.org · llmstxt.org · RFC 9309