Content design starts with evidence – user research, search data, support enquiries, analytics – and uses that evidence to decide what content should exist, in what form, in what order and in what words.
It’s easy to mistake content design for writing with a smarter job title. The distinction matters: a content designer might conclude that the best answer to a problem is a table, a single sentence, a redesigned form or no content at all. Writing is one of the tools available, not the purpose of the work. Content designers usually lead it, and in product teams UX writers often share it, with researchers, service designers and subject experts closely involved.
Why it matters for content
Bad content isn’t necessarily badly written. It can be content that answers a question nobody asked, in a place nobody looks, in a shape nobody can use. Content design attacks that problem at the root by insisting that the need comes before the draft. It tends to produce less content, better placed and easier to maintain, which is usually cheaper as well as kinder.
It does its most valuable work in transactional and service contexts, where someone is trying to apply for something, understand an entitlement, make a decision or complete a task.
How AI changes the picture
AI is very good at producing plausible content quickly, which makes the design question more important. If nobody has established what a page is for and who needs it, AI will happily generate three versions of the wrong thing, all fluent and none necessary.
AI helps with the middle of the process. It can cluster support enquiries, summarise research notes, suggest alternative structures, draft variants for testing and check readability at scale. What it can’t do is decide which user need deserves to exist as content, or take responsibility for the consequences when the answer is wrong. Content design is mostly judgement, and judgement is the part that stays human.
There’s also a newer design consideration: content designed for tasks now needs to survive being summarised by machines. Clear structure, plain language, accurate headings and self-contained answers all help an AI system represent the content faithfully rather than inventively.
Questions people ask
- Is content design just a fashionable name for copywriting?
The two overlap, and content designers borrow plenty from the copywriter’s craft. The difference is where the work starts: content design first decides what should exist at all, in what format and sequence, and only then writes it.
- Do we need content design if we already have UX designers?
Usually yes, and ideally the two work together. Interfaces are mostly language, and language left to the end of a design process tends to arrive as an apology.
- Does content design only apply to government and service sites?
In the UK it was popularised by government digital teams, but the underlying discipline applies anywhere people come to content with a purpose – charities, universities, membership bodies, museums, commercial sites.
- Where does the evidence come from if we have no research budget?
Search queries, site search, support tickets, sales questions and the things staff explain repeatedly on the phone. Most organisations are sitting on more evidence than they think.
Learn more
- Content Design, second edition – the practical book from Sarah Winters and Rachel Edwards on research, planning and delivery.
- Why Content Design Matters: An Interview with Sarah Richards – an interview with Sarah Richards (now Sarah Winters), who set up content design at GDS, about designing around needs rather than polishing requested pages.
- The depleted bookshop – a plea for digital content that can move, delight and surprise as well as inform.