Gauge · Content design
The considerations, in full.
AI is quick on the draft. The voice and the sign-off stay with the practitioner.
A content design team answered three questions about how their work gets made. Gauge read the three together and drafted this. Your own runs the same three questions against your profession, and opens at a private URL you can share.
Where AI fits
First drafts, plain-English rewrites, and reading-age checks: the work you described running through Hemingway by hand. The GOV.UK Service Manual treats content design as a craft of clarity, and a model can take the first pass at clarity quickly, alongside the human work rather than over it.
On the lower-traffic pages you said get less attention than they should, a model can flag the ones drifting from reading age nine before a user does.
Where to hold the line
Voice across the service, and sign-off on user-critical content. You catch voice drift now only when a user complains, often weeks after it goes live. A model can help you spot drift sooner, but the judgement of what the service should sound like stays with the practitioner.
The Service Standard holds the team accountable for the words, not the tool. The senior editor's read before publication is the line, and it stays human.
The answers Gauge read
-
How does a piece of content get to publication?
A content designer drafts from a brief, a second designer reviews, then it goes to a subject-matter expert for factual sign-off and a senior editor for the final read. Most pieces take a week. Urgent updates skip the expert step, which is where mistakes creep in.
-
How does your team handle plain English and accessibility?
We write to reading age nine, run everything through Hemingway, and check against the GOV.UK style guide. Accessibility is alt text, heading structure, and a screen-reader pass on the high-traffic pages. Lower-traffic pages get less attention than they should.
-
When voice goes wrong across the service, how do you spot it?
Usually a user complaint, or a colleague spotting an off-brand phrase in a screenshot. We have no systematic way to catch voice drift across the service. By the time we notice, it has often been live for weeks.
Methodology
How this document was made.
Gauge reads a profession's work for where AI fits, using the GOV.UK Service Manual and the AI Playbook for the UK government as frames, and Cambridge's HCI for AI Systems methodology as the lens. The three questions are written by Chris Leo.
A model read the three answers together and drafted the document in one pass. It is a starting point, not a verdict.
Read it as a working note your team can argue with. Gauge names where AI fits and where to hold the line. Acting on it is the work.