Monday, May 13, 2013

Can an 'open' government site be open if it is poorly designed?

I was chatting with Paul Davis on Twitter recently about the The State Decoded, an open source US-developed platform for accessibly and openly exposing state legislation online (see the platform in use at

He suggested that the tool was effectively a US version of Austlii, which is a repository for Australian federal and state law.

My view was that there were significant differences between the two approaches.

The State Decoded is an open source platform being crowd developed, which anyone can replicate for any jurisdiction. It contains APIs, presents all content as accessible web pages and is nicely designed to be easy for casual users to access.

Austlii, on the other hand, is a closed platform developed by two universities. There's no APIs, much of the content is available only as PDFs and documents, and the design - well, minimalist is possibly the right term, with the site difficult to navigate for all but university professors (who developed it) and lawyers.

When I made this comparison (in brief given Twitter's character limits), Paul said to me...
This made me think a little - do I consider visual design a criteria for openness in government?

And my answer was:
I thought in this post I would expand a little on my view.

For some technically orientated people design can be an afterthought. Their focus is on making a system or machine work as it should, able to take in data and spit out information correctly and quickly.

For these people, design is a 'nice to have' added towards the end of the process, with sites and systems made 'pretty' to appease the communications and marketing people, but is otherwise non-functional.

I've participated in many IT-led 'design' processes, where the focus was on how entities within the system should interact with each other, and the testing focused on 'user-acceptance' - which basically is designed to answer the question 'do the system's features work as intended?'.

In these processes there was little or no consideration regarding the visual appeal of the solution, whether the terminology was understandable to the audience, the search results expected or the navigation logical for non-experts and non-programmers. At best there was some commitment to making the site accessible - however this often meant 'bare bones' lists of text on a white background, rather than using alternative methods  to provide a pleasurable experience for all users.

Of course it is essential that websites and system respond quickly and as intended. However if users don't find them appealing, intelligible or intuitive, they will use them unwillingly, if at all.

I like to compare this to the car market. Originally cars were designed to be functional only - with little in the way of 'frills' to appeal to the public. The hard part was in getting the mechanics to work right and to last and car developers (blacksmiths, bicycle and train makers) weren't concerned about appeal.

Today, however, you'd be hard pressed to find any car maker who doesn't strive for visual perfection as much as for mechanical perfection.

Yes we expect cars to perform flawlessly, but we also expect them to look good. All things being equal (mechanically and safety wise), more attractive cars outsell less attractive cars, people develop more attachment to them, use them more and stick with the brand.

So to with products on supermarket shelves. In many cases people are selecting between products which differ little in their composition (or they don't understand the technical differences), simply choosing on the basis of how the packaging looks and makes them feel. Companies build their brands around their visual and emotional connection with customers, with ingredients a secondary (though still important) consideration.

So it is for software and websites. Well designed software systems and sites attract more use - even where they may be technically inferior (who can tell if a site is a few milliseconds slower than a competitor).

And so it is for open government sites. It is certainly possible to make an open government site with brilliant functionality and the best data - however if it doesn't visually resonate with the audience, if it isn't appealing for them to explore and use, it won't be broadly used.

Governments who seek to be open should recognise that it isn't simply about exposing lots of data, or opening the doors for user participation on a mass scale online. Design must be core to the thinking, how sites are designed, how users interact with the system, the structure of the language and of the navigation.

For openness to succeed in attracting broad interest and active participation from citizens, governments must not only think about what they release, how they release it and how they invite citizens to participate.

They must equally consider the citizen-experience, whether citizens can access information or participate in an intuitive and comfortable way, how citizens feel when using the site - excited, engaged and empowered (for a well-designed site), or frustrated, marginalised and stupid (for a poorly designed site).

Design is important and needs to be involved from the start of the development process. How people should feel when engaging should help drive the features and their operation, rather than trying to 'retroengineer' a clumsy system to meet user needs (a far more expensive and unsatisfying process).

So I stand by my view on open government - a technically open site that is unusable for casual users due to inconsistent, inaccessible & generally poor design isn't open.

Indeed, if a government is only playing lip service to openness (forbid the thought), poor design might be an effective tactic to hide things 'in plain sight', reduce the number of user and 'tick boxes' without revealing anything they are required to publish, but don't want easily found.

So where a government, or agency, releases poorly designed open data or engagement sites (particularly as a second or third version), just as they may release a 'bad news' media release under cover of a major news story, or an old report deep in their site (so they can say it is public even though no-one can find it), citizens really need to consider whether there really is a government commitment, or simply the appearance, of openness and transparency.


  1. +1 to open platforms :)

    Started a conversation today with various country fire authorities - they all publish a list of incidents and so forth, all more or less as georss - it just screams out for a github project to unify the lot!

  2. Craig, I like your definition of openness. If a site's own search facility doesn't produce results after two attempts, I regard it as a failure. So, how about ACT Government websites? For years and years, I have found it necessary to use Google if I want to find anything, even if I have the exact title of a document or topic. Once I did 37 successive searches on an ACT government website and still didn't find what I was seeking!