The gap, as an actual number

What was counted, and on what

Specimen Dental is a practice we invented. It has no patients, no address and no existence outside this website. We built its homepage twice: once in the shape our clinic audit kept finding, and once the way we build. Both versions are ours, both are live pages you can open, and nothing on either of them is taken from anybody's real clinic.

That matters for how you read what follows. Every figure in this article is a measurement of those two pages and of nothing else. It is not an industry average, not a survey result, and not a claim about what your website would do. Two pages we wrote is exactly the weight it carries.

Anaks Labs specimen. Both builds authored by us; the disclosure sits unfolded at the top of the specimen page. What was measured, and on what →

The count

Of the characters a visitor can eventually read on the first build, 716 out of 1,506 were present in the page when it arrived. On the second build, all 2,629 of 2,629 were.

The first number is the interesting one. Slightly under half. A visitor to that page reads the whole thing and notices nothing missing, because nothing is missing from their copy — it fills in while they are still looking at the top of the screen. The other half of it simply was not there at the moment anything else had to work with the page.

The definitions of that row, and of the four other rows in the same table, are published under the table itself rather than being paraphrased here. A count is only worth what its definition is worth, and we would rather send you to the definition than summarise it into something slightly different.

The full table, with every row defined →

The part we argue against ourselves

A comparison like this can be inflated very easily, and the way you inflate it is by putting content into the second version that the first never had and then counting the difference as a gain. We did put content in. So here is the subtraction.

Of the second build's 2,629 characters, 878 are four patient questions and their answers — material the first build did not carry in any form. Take all 878 away and the comparison becomes 1,751 against 716: the same services, the same hours, the same headline, and roughly two and a half times as much of it actually arriving.

We publish the deflated number next to the flattering one because a figure you can check is worth more than a figure that wins an argument. The same instinct is why our sources page shows where we chose the smaller of two available statistics.

Anaks Labs specimen, both builds authored by us. Sample and definitions → · The rule the whole site is held to →

Four things this number is not

Why the gap is invisible from where you sit

Everything you have ever seen of your own website has been the finished version. You open it and the page assembles itself faster than you can register, so the copy you inspect and approve is the complete one. There is no view available to you, in ordinary use, of the earlier and thinner version — and that earlier version is the one that has to answer for the practice when something is summarising you for a patient.

This is the reason the problem survives good taste, good budgets and good agencies. It is not that anybody made a poor decision. It is that the object with the defect is not the object anyone is looking at.

You cannot count your own gap from a phone. You can, however, find the edges of it — the places where something a patient reads plainly is not going to be in the earlier version — and that is what the checklist is for.

Open both specimen builds and scroll them yourself →