Breaking Glass

I must admit that it’s with considerable personal interest that I note that Tom Watson has been given responsibility in his new Cabinet Office job for ‘some of the Government policy towards websites’ – meaning its own web presence and not everyone elses.

Did I tell you that I have responsibility for some of the Government policy towards web sites? I’m having a lot of discussions about them at the moment but in the spirit of not throwing stones at other people’s greenhouses, can you help me with something else?

Will you take a look at the Cabinet Office web site and tell me what you think? How easy is it to navigate? Is it interesting? How can it be improved?

Tom may not want to throw stones at other people’s greenhouse just yet, but I harbour no such compunctions, in fact I’ll quite happily shit in their pot plants as well because the general state of governmental websites has been a particular bête noir of mine for a considerable number of years, so with that in mind I’ve decided to address the big picture rather than Tom’s more specific and limited request for information.

Let’s start at the beginning with the single most important question that Tom should be asking – why do I visit any government website?

That’s the first [set of] question[s] that anyone putting together a website should be asking right at the outset. Why do people visit this site? What is their purpose? What are they looking for from it?

Well let’s start to answer those questions.

I visit government websites for no other reason than the fact that I want information and on almost all occasions what I want is a specific piece of information; a press release, a report, a consultation document, a transcript of a speech, etc.

Whatever it is I’m looking for I almost always want it fast – life’s too short to waste much of it on mooching around idly on government websites in the hope of turn up something interesting and I certain don’t use government websites for their entertainment value. For me they are primarily research tools, a means of gaining information and [hopefully] gaining it quickly.

With that in mind, what I want from government websites are those things that enables me to get to what I’m looking for as quickly as possible; a good accurate search engine, clear and logical navigation through information that is organised in a way that makes sense and which, which I find it, is presented clearly, consistently and in a useable format – which is what I too often don’t get from the government when navigating around it’s overwhelming array of different websites.

Put simply, if Tom wishes to set himself an overriding objective for his role in relation to governmental websites then that objective can be summed up in a mere three words:


There are currently, and have been for a long time, not only far too many individual government websites but within their massed ranks also far too little technical, presentational and organisational consistency between those websites to make using them if not an enjoyable experience then at least a tolerable one.

Let me give you a simple example of what I mean, specifically in regards to technical consistency, with reference to the humble PDF document.

Now I like PDF documents, but I like them for a specific reason – because they enable me to access lengthy [and often expensive] government reports and other documents in full and then read them at my own leisure. What works for me, when it comes to using PDF documents, is the fact that I can download them onto my PC, read them when I like/have time and, if necessary, run out a hardcopy in order to make notes and generally work on/with their content. My requirements when I encounter a PDF document are, therefore, simple and straightforward. I want a simple http or ftp download link, preferably one that indicated the size of the document in question, and nothing more.

There are two main reasons why I only want a download link both of which stem from my preference for using Firefox as my main browser. If I want to view the document straight away I use a PDF download add-on that gives me the option, just from clicking the download link, of deciding how I want to grab the report. I can choose to view it in my browser, download it to my PC without opening it, download it and open it in another programme or even have it converted to HTML on the fly, which is useful if all I want to do is quickly scan a document to see if it one I want. Often, however, I don’t want to read the document immediately because I’m collating information from a range of different sources, so I use a ‘fire and forget’ download manager add-on, in which case I simply right-click the link, select the download manager option and leave it to do the rest while I move immediately on track down the next piece of information I want.

There are some government websites that do, indeed, give me what I want – others don’t. For example, only this morning I needed to get a copy of the Home Office’s new strategy for tackling violent crime. I tracked down the document quickly enough because it’s current enough to be linked from the Home Office’s home page but on getting to download page I was not given a simple download link, rather what I got was a link to a scripted HTML page which either delivers/generates PDF documents or serves, perhaps, as anti-leeching system. The only option I was given was, therefore, that of downloading the document into my browser and then saving it manually from there, which is not that onerous a task as long the connection speed is reasonably high and reliable but it still meant waiting for a few minutes while my browser first loaded the Acrobat plug-in in the background, and then downloaded the whole document into my browser before I could save it. As it was only the one document I needed it wasted only a couple of minutes of my time, however if I’m on a trawl for several different documents and I have to go through the same process for each of them then those wasted minutes soon start to mount up.

Simple things like this matter to me, and if they matter to me then I’d venture that they matter to plenty of others who, like me, use government websites as a source of information in their workplace – it’s a matter of simple efficiency, a plain old vanilla http or ftp download link is a much more efficient way of making PDF documents available for download.

As a general rule – and I hope Tom will forgive me for saying this – the vast bulk of governmental websites should be boring both in terms of their basic presentation and their technical underpinnings, by which I mean they should follow a consistent and uniform ‘look and feel’ in which elements such as the placement of branding (i.e. departmental logos/identifiers), search boxes, navigation, number and arrangement of columns in page layouts, fonts, hyperlinks, document formats, etc should be consistent across them all. One can make legitimate exceptions in certain cases, particularly where there are sites whose purpose is to educate rather than deliver information and which are targeted at a specific audience (which, in the main, will be children and young people) but otherwise, what is, or rather should be, required of functional departmental and sub-departmental websites is consistency and conformity to a common set of standards because, provided that content is then logically and consistently organised as well, what that delivers is an efficient web presence across government – and that’s what I want.

What we’re talking about here is standardisation starting at the base technical level and working upwards.

All government website should, as far as possible, run on the same core platform using the same server operating system, the same web server application, the same server-side and client-side scripting languages and the same database server application – exactly which one(s) is a matter to be determined objectively on a cost/benefit basis the key thing being that once you’ve made your decision – maybe you decide to go with what is the near de facto internet standards, i.e. a UNIX server OS, Apache web servers, PHP and Javascript for scripting and, of course, a standards compliant SQL-based database server platform of which there are several to suit you budgets and likely traffic demands – you implement that platform across the whole of government. Okay, so there will be the odd exception to this general rule, I’ve already mentioned educational sites aimed at young people, which will tend to go for more in the way of ‘bells and whistles’ to grab the attention of their target audience, and there will be some legacy applications that will need to be accommodated using slightly different standards, largely in terms of providing web access to established data systems like the register of companies at Companies House, but for the vast majority of government websites whose primary purpose is simply the delivery of information to the public, one size will fit all.

In fact, in the vast majority of cases, one can go beyond simply standardising the base technical level and take things up a level by adopting a standard content management system across government. One of the most effective ways of ensuring consistency across departments and departmental web sites is to require them to use the same core software application to manage and deliver their website and as with standardising on a single technical platform there are no shortage of options and possibilities open to government, from free, and these days highly sophisticated, open source applications to proprietary commercial systems to the option of commissioning a bespoke system – you evaluate your options, pays your money and you makes your choice; the important thing is that what you end up with a standard system which provides consistency of function and operation across all departments.

Of course, on its own a standard CMS across government is not enough to ensure the kind of consistency I’m looking for, you have go that bit further be standardising what people see and experience when they visit a departmental website. Pretty much all content management systems are, in one form or another, template-based, which is perhaps easiest to explain if you think in terms of something familiar, like WordPress or any other piece of blogging software.

In simplified terms you can functionally divide WordPress up into three constituent parts. You have a database which stores and retrieves the content of a blog; a programme which manages that content, enabling a blogger to write articles and store them in database and visitors to choose which content they want to read and then fetches that information from the database, and then there’s a template which determines how that content looks when it appears in the browser, determining everything from where on the page the name [and other ‘branding’] of the blog appears, to where you find the search box and navigation links to what colour [and size] the text appears in to the reader. This is something that can, yet again, be standardised across the vast majority of government websites in the interests of consistency, efficiency and, also, ease of use – get the standards right and, quite literally, once you’ve used one government website, you’ve used them all, which means you’re not wasting valuable time fannying about trying to figure out how to use the website every time you visit one you haven’t been to before, you just get right with the business of finding the information you’re after.

Looking at just the websites of some of the major government departments (Downing Street, Treasury, Home Office, Ministry of Justice, Foreign Office and, yes, Cabinet Office) the most striking thing about them, when viewed in sequence, is the almost complete lack of consistency and presentational standards – in terms of visual branding there isn’t even anything by way of either text or a logo to indicate that these are all websites that are part of the Government of the United Kingdom, unless one counts the ‘DirectGov’ logo, and even that does not appear on all sites.

Looking individually at each of the home pages of these sites, the two best sites are those of the Ministry of Justice, which is visually the most appealing, and the Treasury both of which sport clean designs and a clear layout. The Foreign Office site is also pretty good, if a bit perfunctory in appearance while the Cabinet Office site is one in which, quite obviously, the notion of design is, at best, a bit of an afterthought but its layout is, at least, functional. The two worst sites are those of Downing Street, which is way too cluttered especially in terms of the five line horizontal menu at the top of the page, and the Home Office site which is, in appearance, little short of foul in its choice of colour scheme and, if anything, badly over-designed. Sorry, but what I’d expect from such a key government department is something altogether more sober and measured in appearance, as befits the serious role it plays in government, not that “Radion orange” abomination – give me a plain rectangular masthead/banner in a good solid blue, police uniform blue, if you like, with a clean silver logo and then you’ve got starting point for a design with a bit of gravitas.

Taken together, however, the impression one gets is – well – little short of being disjointed and unprofessional, that of a collection of unconnected independent online fiefdoms rather than a collection of departments that are all part of the same government. Moving between the sites is often jarring due to the obvious inconsistencies in presentation such that it all ends up looking shabby and not just a little bit shambolic, especially if one starts following some of the links to sub-departments, agencies and projects only to find that some of these are using entirely different designs/layouts to their parent departments.

It should go without saying, but I’m going to say it anyway, that standardisation of the government’s web presence brings with it not only efficiencies for the end users but efficiencies (and cost savings) for the tax payer particular in terms of technical support costs and the costs of developing and deploying departmental and other websites. Poking around the source code of these sites it came as no particular surprise to find that the best designed websites, those of the Ministry of Justice and the Treasury, both use off the shelf commercial content management systems (albeit different ones) – its perhaps a fair bet that these are also likely to be the cheapest sites to run and maintain, all things considered. All that’s need is to take things one stage further and settle on one system, it’s not as if most of these sites are doing anything out of the ordinary that might necessitate the use of unique system, but for things like the petitions system on the Downing Street website – most are just in the business of moving information from A to B.

Ultimately, perhaps the best way to drive home the point I’m making here is with a simple demonstration. Using [preferably] a tabbed browser like Firefox or IE7 (if you absolutely must) open the same six government sites that I looked at – Downing Street, Cabinet Office, Home Office, Ministry of Justice, Foreign Office and Treasury – and take a minute or two just to flip between the site and get a feel for how they look.

Right, now open a new window and take a visit to a major multinational corporation – for my purposes here we’ll use IBM, which means a visit to and if you look at the top of the page you’ll see a blue bar with the words ‘United States [change]’ on it. Click the link [the word ‘change’] and you’ll be taken to a page where you can select any one of ninety-six IBM country websites from across the globe in a variety of languages, from English and German to Russian, Japanese and Korean – pick one at random, preferably one in a language you don’t understand (try Japanese or Korean) and press the ‘go’ button – it’s the round blue one with the arrow.

Notice something interesting?

It the same site – in a different language of course and the pictorial advert is a little different but it’s the same look, feel and layout.

Now, let’s head back to that site selection screen, but without using the ‘back’ button on the browser – you need to find the same link you used to get there from the US site…

Found it?

Of course you have – it’s in the exact same place so even if you can’t read Japanese you know that the bit of white text in square brackets on the blue bar at the top of the screen is what you need to get back to the country selection page.

Okay, now try a different site – Sweden.

Ah… not that looks little a bit different doesn’t it – it’s a lot plainer in terms of design, with none of nice rounded corners, and a bit smaller in size but otherwise the layout is pretty much the same. The menus are in the same place (as is the link for getting back to the country selection page), as is the search box and quite a few other layout elements (if you haven’t noticed, the ‘contact’ link is on the black bar at the bottom on all of IBM’s sites regardless of the language in use.

Now try a couple of other sites is different languages… what you’ll find is they all follow the same basic layout, some have the fancier rounded design, others the plainer one (like Sweden) but otherwise the general look and feel of each site is consistent and its consistent across 96 different countries and around 30 languages, including one (Hebrew) that reads from right to left and not for left to right.

Now do the same thing with these that you did with the government’s site – open five or six in different tabs in the same browser. Which creates the better overall impression, the government’s hotchpotch of different designs and layouts or IBM’s consistency of layout?

It’s IBM, of course – in fact once you’re familiar with the layout I dare you may feel fairly confident about finding one or two of the standard links on one of IBM’s non-English language sites without too much difficulty or trial and error even if you have not conception of what the text in front of you actually says. The ‘change’ button (obviously) but if you’ve paid attention you’d also have a pretty good shot at finding the ‘contact’ link and maybe even the link for ‘downloads and support’ as well, if not first time then after check no more than a couple of options.

If I gave you a list of 96 different governmental websites from the UK, could you honestly feel as confident about finding the ‘contact us’ link straight away without having to hunt around for it?

Point taken?

4 thoughts on “Breaking Glass

  1. You did miss one important thing. W3C compliance. Taking a little look at the CSS validation for the Cabinet Office’s site alone, it’s very poor to say the least, which probably explains why the top bar menu appears out of alignment on my browser.

  2. It’s simpler than that; if you are on a government Web site, you’re either there to do something, usually a transaction of some kind, or to find something out.

    The first category needs to be clearly signposted by service, not by organisation (which, to be fair, seems to be what they’re doing with Directgov/Businesslink), and further, everything they provide should be available as a Web services/XML/SOAP/Whatever API.

    The second category primarily needs to be search-friendly, as people will mostly arrive there from search engines, and probably will be not sure what they’re after, hence they will search and browse. Once they find it they will want to get back to it, so stable and humane URLs are important.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.