Why we scan a business before we quote it

Most proposals for small-business web and marketing work are written from a conversation. The owner describes the problem, the agency describes a package, and a number appears at the bottom of a PDF. Nobody has looked at the thing being fixed. The number is a guess dressed as a quote.

We do it the other way round. Before we scope anything we run a scan of what already exists — the public site, the paths a visitor can take through it, the ways an interested person can actually make contact, and what is measurable from the outside. The scope then answers something observed rather than something assumed.

What a scan actually looks at

The scan reads the site the way a machine reads it: no JavaScript execution, no assumptions, just what the pages return. It scores a set of capability categories, and each score carries the evidence and the formula that produced it, so any number can be recomputed by hand.

• Website — does a structurally complete, reachable site exist across the paths people try?
• SEO — is the site legible to a search engine: title, description, canonical URL, one heading, structured data?
• Branding — is the brand presented consistently, and is there an organization record a machine can read?
• Trust — HTTPS, a published contact route, policy pages, security headers.
• Accessibility — alt text on images, a declared language, a responsive viewport.
• Content — is there enough substance on the page to sell from, and any signal that it is maintained?
• Lead capture — can an interested visitor become an enquiry without hunting?
• Customer journey — is there a path from interest to enquiry: services, pricing, about, internal links?
• Performance — how much weight the page ships.
• Social presence — are there real profiles linked from the site?

What the scan refuses to do

A category with no supporting evidence is reported as unassessable, not as zero. This distinction matters more than it sounds. Absence of evidence is not evidence of absence: if nothing on a site says anything about how the business runs internally, the honest output is “not measurable from here”, not a bad score.

The same rule applies to whole domains. Delivery — how work actually gets done once a client signs — leaves no trace on a public website. So a scan does not score it. We ask about it instead, in the conversation where it belongs, rather than inventing a number that would be indistinguishable from a measured one the moment it was written down.

Why the order matters to you

Diagnosing first changes what you are buying. Instead of a package chosen from a menu, the work is a response to specific, named gaps — and you can see the evidence for each one. If the scan finds that the site has no enquiry form on any page, that is a fact about your site, not an opinion about your marketing. Fixing it is measurable. So is not fixing it.

It also changes what we can honestly promise. We will not quote for a rebrand because rebrands are what we sell; we will quote for a rebrand when the scan shows the brand is presented four different ways across the site and there is no organization record at all. Those are different conversations, and only one of them respects your budget.

What you get from the scan itself

A scan produces a read on each category with the evidence attached, a list of observed weaknesses and risks in priority order, and a statement of what could not be assessed and why. It is useful on its own even if you never hire anyone: most of the highest-priority findings on a small-business site are half-day fixes that nobody had written down.

That is the point. The scan is a diagnosis, and a diagnosis you can act on without us is a more honest instrument than one that only ever concludes that you need us.