https://gds.blog.gov.uk/2012/11/05/tools-over-content/

Tools over Content

I’ve talked before about various aspects of how we build software at GDS, and especially for GOV.UK, but I’d like to take a step back and about one of our early design rules: “tools over content”.

Many of the needs we’re addressing are complicated. If the team building a website only have a generalised ‘content management system’ it can become very difficult to provide people with exactly the information relevant to their situation. As an expectant parent you’re likely to want to know exactly what your eligibility is for leave and benefits. The full details of your entitlements can be very complex depending on due dates, when you started your current job, your employer’s policies and so on. Most users don’t want to consume and interpret all that information themselves, but if the content management tools only allow us to publish text then that’s what users will end up reading.

In building GOV.UK (and establishing GDS) we knew we had to do better than that. We had to do the hard work to make it simple. What is complex as a piece of prose can often be simplified.  Asking a few questions can take someone to a very specific and tailored answer.  It can also take them directly into the transaction they need to complete. By asking the right questions we can take you straight to what you need to do rather than making you wade through hundreds of words to determine exactly what content applies to you. Even where the answer involves a fair amount of reading we can give people tools to penetrate, navigate and apply all that content. An example of this is our maternity leave planner which will only require most people to enter a couple of dates to find out what leave they can take.

Greater than the sum of our parts

The secret is to have a cross-functional (or cross-disciplinary) team. You need people who understand the needs, the users, the available tools and the creative possibilities. Those people need to be able to challenge one another and collectively refine the solution. If they can do that they will usually come up with something much better than any of them could have achieved alone. That’s been a big part of what we’ve been looking for when we’ve been hiring.

Each of the “formats” we’ve used to shape GOV.UK have emerged out of that cross-disciplinary process. A first iteration of any format or tool is specific to a single need. Future iterations and newly understood needs may identify that format only suits one need, or that there’s scope to clarify and develop it to serve many more needs.

As formats have become more clearly defined we’ve found that content designers drive the formats that are most clearly prose, business analysts drive the smart answers, developers wrangle the more awkward data-driven content and so on. In each and every case, however, it takes a mix of skills to identify what format best solves a need, how we continually refine the formats and when exactly we need to build specific tools.

8 comments

  1. alanmaddrell

    The smart answer format has been hugely valuable in allowing the content designers to get rid of the “if this, then that” (as James says). It’s great to have the tools to enable us to do this.

    For example, this started life as a 90-odd page PDF:

    https://www.gov.uk/legal-right-to-work-in-the-uk

    Content designers worked on the logic and text, analysts and developers refined and built the interaction and fact checkers helped us get it right.

    There are more to come: watch this space.

    Reply
  2. tomrup

    Great stuff, and I can empathise with the amount of hard work that goes into distilling complex law into a few simple questions.

    Interactive tools on Government websites didn’t appear for the first time last month. Both Directgov and Businesslink had plenty of tools to help users avoid wading through pages of regulation. BL, for example, had more than 40 such tools including very popular holiday and maternitypaternity leave entitlement planners.

    Hopefully at least some of the logic behind these tools will be reused as a starting point for new smart answers. As far as I know the law is in most cases the same…

    I’d love to know your approach to securing approval from policy owners for the tool outputs – always a challenge where the maths is not necessarily linear, and the variables abundant.

    Reply
    • James Stewart

      Good point – we’re definitely not the first to tread this path and have learned a lot from what came before us.

      One of my colleagues is hoping to write up some more concrete case studies that should give more insight into the way we’ve worked with policy owners.

      Reply
  3. Tim Manning

    A bit of a dangerous title (!), but I’m pleased that good design practices are being applied right at the centre of government. Unfortunately, more often it’s “project management over design”, with little ‘design’ being in evidence; leading to complex and poorly designed solutions.

    Hopefully, this resurgence of design can be probegated further, into other corners of the public sector. On-line transactional services will be the opportunity and challenge!

    Reply
    • James Stewart

      Dangerous? Maybe! Using the phrase internally is definitely meant to provoke us to think carefully whenever we set out to add anything to GOV.UK

      For the transactional side, today’s launch of the Government Digital Strategy is another important step in that direction and we’re really looking forward to working more with colleagues elsewhere in the public sector.

      Reply
      • Tim Manning

        Sorry, a bit cryptic. Only dangerous if taken out of context.

        In the wider world of service design, the over emphasis of tools has been much criticised, e.g. technology fads such as CRM, use of lean tools (developed for manufacturing), etc. The key is that tools are pulled into the design problem faced – as here, not used simply because they exist!

        Reply
  4. Jon Lawson

    I like the idea of having a unified approach that focuses on providing what the end-user needs as simply as possible, but when I followed the link to the example provided (maternity leave calculator), I was rather unimpressed. First there is an unneccessary “Start now” button to click, then you have to select a date from three dropdowns that includes dates more than 9 months in the future, and well into the past, as well as allowing you to select imaginary dates (31st Feb anybody?). After that you are presented with another set of dropdowns asking you for your preferred maternity leave start date, with the same unlimited date range. Only after you have selected an invalid date does it tell you what the rules are (requiring sitting down with a calendar to work out).
    Far better if the first screen just presented you with a calendar of the next 9-10 months and asked you to click on your due date. The next screen could show the same calendar with the allowed start dates highlighted and clickable. Much clearer, and the user would get to the same point with only 2 clicks rather than 9!
    It seems as though the “Do the hard work to make it simple” principle was only partly applied to this service. That said, it is brilliant that a real effort is being made to bring provide these services digitally in a way that makes sense to the people using them, rather than the civil servants and technologists who administer them.

    Reply
    • James Stewart

      You’re right that there’s definitely more that we can do to improve numerous aspects of the site! Now that the launch is complete we’re working through the data about how the site’s performing and will be continuing to iterate a lot of the content and tools, maternity and paternity leave calculators among them.

      Reply

Leave a comment