Articles and talks

Setting Standards for Website Design: New government framework a missed opportunity

by Louise Ferguson 2003

[The following article first appeared in eGov Monitor - a publication directed at a non-specialist audience - on 18 August 2003.
It was republished in Usability News on 19 August 2003.
It now appears in the Usability Views Usability Heretics collection.]

The number of public sector websites has mushroomed in recent years, but governments around the world have been slow to provide design guidance for them. Consequently, approaches to this channel have often been well-meaning but ill-considered.

Over the last year, the Office of the E-Envoy has published a stream of website guidance, addressed to a range of audiences. The latest is the Quality Framework for UK Government Website Design: Usability issues for government websites.

When the word 'design' is mentioned, many people think of graphic design, something 'superficial'. But designing websites involves a complex bundle of interrelated activities with decisions about content, terminology and structure, as much as about surface, or the 'interface'.

Discovering who the website's potential users are, finding out their requirements, using this to decide what information to include on the site, how to structure it and what to call it, and designing navigation paths through it that are as clear as possible, are all part of design. These days, it is also essential to ensure information is accessible to as many different people as possible.

Many of the best practices regarding such processes are now reflected in ISO standards, which are not particularly readable. We need others to interpret those standards for us, putting them into usable form.

Usability - there are many definitions of it, but a simple one might be the ease with which people are able to use something - is something relevant to every part of our lives, but rarely mentioned in public. Any incursion of government into usability has to be applauded, while to see a substantial government report addressing the subject is, for any usability professional, a rare and wonderful thing.

So how does the latest publication shape up? Does it reflect current thinking on the design and usability of websites, and will people find it useful? Indeed, will they use it? Ultimately, will it help improve our public sector websites? In general, the Framework contains much good advice, and its separation of usability from accessibility is welcome too - they are two quite separate ideas.

In principle, the Framework sets out government support for user-centred design for websites: putting the person - not the organisation, or the design team - at the centre of the design process, supporting ISO 13407.

It does not, however, discuss how government websites might be different to other kinds of websites, or address their specific challenges. There is an excessive focus on third party 'agencies' and discrete web projects, rather than a grappling with more difficult 'joined-up government' challenges. There is no presentation of best practice (for navigation, for example). All in all, I find the Framework a good idea in principle, but poorly executed.

It suffers from confusion and disorganisation, together with poor writing, while some sections can only be described as misleading. It is clearly a document written by a committee, with little integration and too much cut and paste.

I wondered whether I was the only usability professional who on the one hand welcomed the government embracing and promoting usability, but on the other saw gaping holes in the emperor's new clothes. I decided to consult colleagues in industry and academia.

The comments I received confirmed my views. While we can award the new Framework a definite 'A' for effort', it should be considered a point of departure, a version 0.1, rather than any kind of solid basis for future public sector website design.

Speaking on behalf of the UK Usability Professionals' Association, president Giles Colborne said, 'The E-Envoy's office has been pushing usability up the eGovernment agenda - and the framework is another welcome example of its advocacy. We would like to see the Framework expressed in a simpler - more usable - form so that it is understood as widely as possible. We hope to support future iterations with feedback and suggestions from our members.'

'It doesn't make for clear, easy or compelling reading and lacks visuals and schematics,' commented one practitioner who preferred to remain anonymous. 'What's more, it suffers from almost unstructured content and is full of largely unsubstantiated assertions, odd references - a really mediocre 'committee' production.'

Gerred Blyth of agency Amberlight was equally critical: 'Many Web managers will be left utterly confused on a sea of acronyms, blurry definitions, jargon and seemingly unrelated methods, models and standards. It should have been an accessible round trip of usability issues and approaches - instead we have a muddle of thoughts and proclamations.'

'It's encouraging that they've done something, and I feel the government is very much to be applauded for taking usability seriously,' said Geraldine Fitzpatrick, formerly of agency Sapient and now senior lecturer at Sussex University. 'But it's a bit of a missed opportunity.'

And for a document about making the best use of new media, it makes surprisingly little use of online resources. Instead, we have many offline references to academic papers; what online resources are included are often out-of-date.

'The referencing of the 'jthom' usability toolkit was just plain lazy', said Jon Dodd of consultancy Bunnyfoot, referring to a page-long list of repeated references to the same website.

Perhaps the ultimate irony is a document focussed on the subject of usability that is not at all usable.

It's not only among consultancies that dissatisfaction can be found. Ann Light, editor of Usability News, is forthright in her criticisms.

'While I am relieved that the Government recognises the need for some contact with the end-users in the design process', she said, 'the document's detail reverses any confidence I was beginning to feel in their understanding of how to manage this involvement.'

A surprising aspect of the Framework is the conflation of user testing - putting people in front of test materials - and the idea of human-centred design, according to Fitzpatrick. Usability is not just about testing, she pointed out. 'I think the biggest hole is not taking the opportunity to discuss iterative user-centred methods - there seems to be this 'do it and throw it over the wall' approach.'

The benefit of ISO 13407, supposedly at the centre of the Framework, emerges when it is used to guide an iterative development process, repeated cycles in a design loop that move us closer and closer to a better design.

But the Framework somehow manages to turn the circular idea of iteration into something linear.

'I don't think anyone is stressing iteration enough - and here it's demoted to an appendix,' said Phil Barrett of agency Flow Interactive. 'For user-centred design to really work, organisations have to change the way they operate.' And that's not easy in either business or government.

And what about the extent to which the Framework reflects current professional practice?

'I feel it very much reflects one company's view,' said Fitzpatrick, 'and is written from the point of view of a usability company generating work for itself.' In recent years professional bodies have moved away from seeing usability as something done by third party agencies, towards a view of embedding user-centred iterative processes within the organisation. 'It strikes me as reflecting very old school thinking, a document written by people out of touch with current practice.'

Usability and user experience are not just about providing a service - the idea is to move towards a situation where the organisation begins to modify its own internal processes to a more user- or human-centred approach to its work.

One section of the Framework has proved particularly controversial is that discussing the optimum number of people for user testing. The advice is that: 'Once groups of more than eight users are used, the law of diminishing returns applies. Other companies prefer to observe one user at a time carrying out tasks and to ask them structured questions about their experience.' Light was particularly critical of this approach.

'The advice - seemingly from the industry - that multiple users be tested simultaneously has not come from any theorist or practitioner that I know of,' she said, 'and could lead novices into treacherous waters. I have no idea where they got the idea that this was good practice, and I will not be alone in raising my eyebrows.'

In general, the choice made of resources referred to is poor, and there are some surprising omissions, including the widely-consulted UsabilityNet, a website funded by the European Union to promote usability and user-centered design.

The Framework is a document about the Web. So why not link to online resources? In fact, why not create an online resource that supports the advice given and that can be easily updated?

Perhaps the document reveals more about its devisers than it tells us about usability practice.

'As a practitioner, I find the document a good insight into the current thinking of the contributors,' said Blyth. 'However, it does not appear to be useful, or usable to the intended audience. And I daresay, as yet another anonymous E-Envoy offering, it will not be used.'

But there is a way in which the document is being used that does raise concerns. In a proposal for a government department contract last year, the following statement was made to bidders:

'The bidder is strongly recommended to use the emerging 'Quality Framework for UK Government Website Design' produced by the Office of the E-Envoy and shall state any deviations from, or issues with, this approach.'

We have it seems a potentially very powerful document, but an excessively flawed one at present.

Read more articles on this website
Check out shorter pieces on the blog