Monday, September 13, 2010

What does it cost to build and run a government website?

The Guardian reported in July that the UK government has released details on the costs of developing, staffing and hosting their major government websites.

The data includes web traffic, accessibility and user opinions on the websites.

This type of data is very useful when modeling the costs of developing and operating government sites, allowing agencies to more accurately forecast costs and staffing needs. It allows agencies to compare their web operations with other agencies, providing a view on who is most - and least - efficient.

The approach also allows hard-working, poorly resourced and funded web teams to more effectively argue for a greater share of the agency pie.

I would love to have such data available here in Austraia - down to being able to derive a total cost per visit (which for UK sites ranges from 1 pence up to 9.78 pounds - see the Google spreadsheet below). It would significantly assist web teams and agencies in their planning and activities.

The UK website data can be downloaded here.

Or see the data visualised (using IBM ManyEyes) and a Google spreadsheed of the costs below.

Download the full list as a spreadsheet


  1. Agree - would love to have this sort of data published for Australian govt. websites - and also US govt. US cost data would be another interesting point of comparison with how much we are spending here.

  2. On the one hand - it's always great to see more data being published. BUT - I have two problems with this.
    The first is that the technical complexity of one "website" compared to another is very hard to quantify. Imagine trying to compare the cost of building the BoM site with PM&C.
    The second problem is that I always worry that the focus on the cost of technical development will overshadow what ought to be much greater cost/investment - the effort in maintaining the currency of the site once it's up. Anybody budgeting for a new site needs to keep it at the forefront of their mind that.... well, a website is like a puppy: it isn't just for Christmas. It's a lifelong commitment.

  3. I believe a more useful (and more difficult) measurement methodology would focus on 'value' rather than input costs.

    When focussing on input costs alone - it can lead to situations where a $100 per click site that saves a life is deemed to be worth less than a 1c per click site that distributes transport brochures.