How not to spend £180,000 on a website

Man with money

Defra has been widely criticised for spending £180,000 on redesigning their website – because it looked ‘too brown’.

Now it’s easy to poke fun here, but as web designers we’re interested how that money was spent and why. I mean, the money didn’t just go on a colour change… right?

Here’s the old site:

Defra-old

And the new one:

Defra-new

Certainly, there is a lot less brown. And it is a lot easier to use.

However:

We found a lot of this odd – normally, we can’t do a 10 page website for a small council without more stringent quality control. I don’t believe we’ve ever seen a public sector site that doesn’t insist on W3C compliance.

So where did the money go?

£32,000 went on “web design and templates” and £40,000 on a “usability assessment”. Which leaves £108,000 for “other”, which would usually split between buying software, hardware, photography, copywriting and manually copying existing content from the old site.

It doesn’t include the cost of Defra’s own staff by the way – that’s extra.

The Defra website is big – about 9,000 pages, and 20,000 documents. However it isn’t actually that complex – most of the pages are just text, and they appear to have been copied over using an automated process (if not, they should have been).

The design isn’t bad. It’s mediocre, which is almost expected from big projects that have to appease zillions of stakeholders. A prettier design would be easy to create, but largely pointless.

No, what really captures the imagination here is just how so much money was spent. £180,000 is enough to pay 7 people £25k each – and have £5k left for tea and biscuits. That’s 7 people on a salary of £50k, working solidly for six months. While the site is better, it’s simply no way near good enough.

We don’t know if the money went into bureaucracy, launch parties or fat margins, but it could have been spent better. Defra appear to have paid out for a Rolls Royce and bought a Morris Minor.

  • roger devlin

    there are hundreds of web sites owned and/or paid for by government departments. lots of branding/vanity/unmanaged information and misinformation. ask yourself why the bureaucrates just love to waste public money on thier own web sites and why are they resisting transformation of government web sites into the direct gov melange

  • http://www.churchdigitalagency.com Darryll Bayes

    Ok. Something has definitely gone wrong here. A redesign of a website is usually driven by a need, whether this is an internal view to ‘spruce up the site’, or more of a business critical need that has come from the top. Either way, this would (should) have been tendered out and a joint board decision would have been made to appoint the right agency for the job. This decision should have been based on experience and quality of work from the agency.

    If they spent £32k on Design and Templates alone, then someone has got a big smile on their face. They should have paid for the Information Architecture to be done before even looking at the visual design (maybe this is where some of the £108k went?).

    £40k for a usability study is not unreasonable for a site of this nature. However, doing the study is one thing… but doing something about the the outcome is more important. (again maybe some of the £108k).

    And so… looking at what they have got for their money – I really don’t know where to start with what is wrong with it, but here is a brief list…

    VISUALLY
    • Positioning of main components are not intutive

    • Main and sub navigation is lacking in clarity

    • The use of colour is… well… subjective, but I am sure it wouldn’t adhere to their corporate online guidelines (they have got these, right?)

    • It is lacking in any visual impact to engage with the user to keep them even vaguely interested in what they are trying to accomplish.

    • Web design doesn’t mean lots of over the top effects with ‘apple-esque’ styling and lavish buttons, but can be very effective with good use of space, colour, fonts and lines… all of which can be achieved through CSS… but sadly none of which has made its way into the site.

    CONTENT AND ARCHITECTURE
    • With a government organisation of this size, they will have different audiences visiting the site, and should have focussed on thr main aims of the site to address the primary audiences well. I do not see a clear definition of how the site guides the user down a well thought out path to the information they are looking for.

    • The page content itself comes across as very unfriendly and no consistent tone of voice. Ok, it has probably been written by quite a number of different people and departments, but a copy guide that outlines how best to write copy for the web that also has a guide to how you speak to your audience goes a long way to making sure it is ‘one-voice’.

    DEVELOPMENT
    • It has been quite a while since I have seen a site this size built in HTML! An HTML site is perfectly fine for a ‘brochure’ site that has maybe 10 pages… it’s quick, it’s cheap, it’s easy. But it is not a language that should be adopted for a site of this magnitude. It is not very flexible, it does not allow for dynamic content very well, it does not really have much capability for future developmental features and is just, well, a bit basic.

    • Taking a quick look at the code (which is cluttered with unnecesary code and spacing) it is clear that it is not compliant in any way, and of course running it through SiteRay you will see the literally thousands of issues that it throws up.

    • SEO – well… maybe they have got something right (kind of), I see they are using Dublin Core Metadata Initiative to describe each page… or rather they are attempting to. Most of the pages I checked were left half completed of the data, which, to be honest I can’t blame them!

    • As for the other basic SEO needs.. there is a lack of descriptive TITLE tags on most of the pages and other basic META data (not that search engines really pay much attention to that these days, but at least they could have tried!

    • CMS – I am going to hazard a guess at Adobe Contribute?? Again, this is fine for small brochure sites… but a 9,000 page site? Really?

    And so.. I could go on..

    It doesn’t really bother me that they have spent £180k on the project – to be honest it sounds about right, maybe even a bit on the low side depending on the amount of copywriting.

    But the thing that doesn’t sit right is what they have got for their money.

    If they wanted to do it again, their budget should have broken down similar to:

    £10k Defining website goals and objectives
    £25k Information Architecture (Including wireframing and sitemap)
    £8k Usability testing of wireframe
    £5k – £10k Changes to Information Architecture
    £10k Template designs
    £20k Additional page designs
    £12k Content harvesting/assessing
    £20k Copywriting/Copyediting
    £40k – £60k Development (depending on solution, possibly Sharepoint or PHP with custom CMS)
    £10k Usability testing
    £10k Amends based on testing
    £15k Project Management

    It is when public or indeed private companies get it so wrong with big budgets that they catch the eye and get exposed as to just how much of a balls up they really made.

    Still, there’s always next time.