The history of content design in the UK government

This is the first blog post in a 4 part series on how GOV.UK and content design happened at Government Digital Service.

The term ‘content design’ gets a lot of discussion. Mostly, is it an accurate term for what we, content people, do?

I wanted to tell you how it came about for the British government. This is the first of 4 posts about the GOV.UK beta from a content (and my) perspective.

This first one is background but to understand the content foundations of the Government Digital Service (GDS), you’ll need to understand the environment they came from.

Massive caveat: my memory is imperfect. I will have forgotten things and people. I’m also skipping over loads to save space. No offence is intended.

UK government, 2003

All government departments are merrily publishing content to their many sites. At one point, it was estimated government maintained over 3,500 government websites.

Obviously, what you need in that situation is another website.

This one was called ‘the storefront’ (as in storefront of government). It was to take over from UKOnline, a portal site to help you find your way around the 3,500 government websites.

Eventually, storefront became Directgov, which launched in April 2004. This big orange behemoth was the government’s site for all citizen-facing information.

Departments would then work with the Directgov central team to (mostly) duplicate the information that was already on their own sites.

The ‘right’ publishing model

In government, staff liked writers to write and sub-editors to edit and no-one was allowed to disagree with policy or legal who signed off content. The person’s skillset was firmly in their title. Easy. Everyone knew where they were.

At that time, the process in some departments was:

  • someone writes (not necessarily anyone trained in writing, web writing, copywriting, journalism or anything like that),
  • it goes to through an approval process with legal or policy people having the final say (one dept had a 27-point approval process. It took up to 3 months for each piece of info to be published),
  • anyone in the departmental content team who suggested changes were generally overruled by legal,
  • content went to Directgov,
  • the Directgov editors suggested changes (often the same ones the departmental content editors did),
  • they are overruled,
  • content was published.

Not an ideal situation. For a lot of departments, it went on like that for years. It was inefficient, frustrating for most of the people in the whole publishing process and it wasn’t good for the user.

One of the most important things to remember, from this content-biased view of this history, is that Directgov was paid for by the departments. They all had to give money each year for its continued survival. It was quite a gruelling process I understand (I had nothing to do with it so can’t comment directly). The departments knew this. They used it. They said they would refuse funding if we didn’t do it their way.

I am not exaggerating when I say I had conversations with people in departments who said: ‘Sarah, we’ll just go over your head and get our way anyway so you might as well give in now’.


I led a team in Directgov central’s publishing team. We were working on convergence, part of a program to shut 551 sites. It was the first major digital cull for government.

(Full list of sites to get the axe is listed on the BBC.)

We had:

  • a weak mandate (departments had to shut their site – nothing was said about quality or quantity),
  • no back up (confrontation was not a strong point for many in government at the time),
  • bad relationships (departments had previously got their way by simply refusing. Why change now?).

The departments and agencies who were moving their content, for the most part, didn’t want to move. And it was hardly surprising. We had a bunch of immovable templates, an unwieldy content management system (CMS) and machines that took 15 mins just to start up in the morning. It took an average of 2-3 hours to put a 750-word article into the CMS. As a government, we’d also been in-fighting for years over jargon and usability. Not an ideal starting position.

My team had to work with the government departments to get their content across to Directgov. I can honestly say, I didn’t learn much about content usability in the whole process. I was reading a lot on a personal level but at work, I learnt the most about human behaviour, which included:

  1. people loved their site with all their heart,
  2. many people who had been doing things for at least 5 years didn’t want to change,
  3. most people thought if you want to test anything, you were clearly rubbish at your job,
  4. being passive-aggressive can be an art form.

So we found ways around it. Some of the departments and agencies were amazing to work with. Sometimes, we in central, didn’t behave well. Eventually, we got through it.

These are the main lessons I learned in trying to remove thousands of pages:

1. Departmental editors were human

At Directgov central, we often wondered why the departmental editors would allow such shocking content.

In convergence, I met a couple of the people they were working with.

(The process was Directgov central would only work with department content teams, not any of the agencies etc that were behind them. We weren’t allowed to talk to the experts.)

At Directgov, we had a small team of user researchers and we were learning. The rest of government had varying levels of skills and some had never heard of user research. I thought we were having a tough time; some of these people were having a much worse time of it. Especially those who didn’t have training or digital content skills, and they were caught in the middle.

2. Cake and biscuits are very important

In government at that time, we weren’t allowed to order any kind of refreshment (tea/biscuits etc) unless we were in a meeting for over 4 hours, over lunch time and with 10 external people (or more). So I invited people to tea and cake and paid for it. (Convergence cost me a fortune.) It was weird. People were on their guard. Some thought I was ‘buttering them up’ (I was). But it was a gesture and half the time, it worked. It at least got them to listen – and that was the start.

3. Under the radar is a legitimate place to be

I absolutely shouldn’t be saying this. But it’s true, so I will. If my lovely team and I played by all the rules, we never would have got anything done. We used personal relationships, worked round people, went over them or under them. Whatever it took.

People across government got the job done and many sites were shut down. Hurrah.

Was it all shiny when it ended? Nope. Was it brilliant? Nope. Was it better than what we had before? Yes. And there is no way GOV.UK could have been published in the time it was if Directgov wasn’t there.

But that part of the story is for next week. This week, I’d like you to take away that, for a lot of government, research was new or seen as a failure, silos were rife and more like heavily defended fiefdoms and there was very, very little trust.

Read the next part of the story with GOV.UK beta: the beginning.

Sign up to our newsletter

Get content design insights sent straight to your inbox.

  • Choose what information you get: (required)