This is the second in a series of posts about how I am taking forward objective two of the single domain project: specifically the “private beta test of a shared ‘corporate’ publishing platform aimed at replacing most of the activity currently hosted on numerous departmental publishing environments”. By corporate, we mean the parts of the government web estate - currently accessed through separate domain names such as bis.gov.uk, dh.gov.uk and number10.gov.uk - which describe the aims and purpose of government's various organisations, explain in detail the work they do, and provide information about how they are doing it for transparency and accountability.
This post is about the needs the beta version of the corporate publishing platform will be designed to meet, and how I have gone about answering two deceptively simple questions:
- Why do government organisations have corporate websites? and,
- What do people want when they visit them?
But before we dive into why and what, I first need to give you sense of how many and who.
How many people are using these sites?
Lots. The average monthly unique visitors to central government's current corporate sites range from 3,000 (lowest) to 4.3 million (highest). Crudely, this puts the total audience at about 12 million people - not counting the websites of agencies and other arms-length bodies, many of who dwarf their parent departments (think, for example, of the Met Office). A high proportion of visits are from repeat users - about 40%. Users are highly engaged, clicking an average of 4 times per visit. (Hat tip to Adam Bailin for these numbers).
Who are these people?
Everyone. Visitors to government's corporate sites are best described as “people who are professionally or personally interested in the work or workings of government” - more of a mode than a demographic. There’s unsurprisingly a bias towards the professionally interested: traffic falls away in the evenings and weekends.
While broad grouping is possible, visitors defy neat categorisation. Surveys of users of these sites tend to throw up a lot of “others”. To give you a flavour, here's how 10 users (picked entirely at random) described themselves in recent surveys by the Ministry of Justice and Department for Business, Innovation and Skills.
Ministry of Justice
- Neighbourhood watch development officer
- Forensic scientist
- Partner of prisoner
- Claimant in insolvency case
Department for Business, Innovation and Skills
- HR advisor in the NHS
- Further education compliance officer
- Board member of a third sector organisation
- Local government officer
Making sense of the long tail of needs
With such a variety of people interested in such a variety of subjects, it was clear early on I'd have to take a more coarse-grained approach to identifying user needs than has been done for the citizen-facing part of the domain. A fine-grain study à la (brilliant) Needotron 5000 would have bogged me down in analysis paralysis and made delivery by early 2012 impossible.
In any case, coarse granularity makes good sense here. My goal in this exercise was to establish the needs that would be met by formats for content as opposed to the content itself. My immediate job is to create the right shelves; later I will work with Departments to stock them, and during that time we will need to think carefully about the why and who for for every item we add (and subsequently update or remove).
I also chose - unsurprisingly given my background - to study government's motives for their corporate websites, as a check on how compatible they are with what users expect, and because pragmatically I need to meet both sets of needs if this thing is going to be a hit.
So I set about gathering just enough data to meet those objectives and no more, asking Whitehall’s webbies to send me their:
- Search keyword data, web analytics, user surveys and research insights (everything they've got on who comes to their sites and what they are looking for)
- Website strategy documents, propositions and KPIs (any documented thinking from the departments about the goals, audiences and success measures for their sites)
I got 70 documents back, speed-read the lot, and distilled them into separate lists of user needs and government goals. I ran the results past digital leaders in departments for a sense check and arranged for a second pair of eyes (the fabulous Mo Brooks in COI) to clean up the duplicates and group them into themes.
Here are the raw results of all that work: download the Excel file or view it on Scribd.
I hesitate to put any spin on the results, but would like to share just three observations that jumped out at me during the process. Your mileage may vary.
- Users want the latest stuff. As I flicked through search keywords, most viewed pages and stated reasons for visiting, it felt overwhelmingly the case that users of corporate sites are looking for information about recent announcements.
- They also want really specific stuff that's current (as distinct from recent), and are looking for it by generic search terms (e.g. from DECC: “energy efficiency”), by document titles ("climate change act 2008”) or named initiatives (“renewable heat incentive”).
- There's a two-way relationship between users and government officials which corporate sites could do much more to support. Where user and government needs align neatly (e.g. understand/explain, influence/consult, stay informed/announce) and where they don't (hold to account/get positive PR, lobby/campaign), there is a tacit negotiation and a desire to influence behaviour in both directions. If the single domain were to support this better, it would need to become more of a conduit for policymakers and their audiences to engage in dialogue and open exchange of information, to foster a better understanding of each other’s wants and needs.
The next step with this work is to map needs to formats - which we've started on now - and to keep iterating these lists as we develop and test the beta product with users.
Thoughts? Questions? Needs you think we've missed? All comments welcome as ever.
Photo credit: Alice Bartlett under Creative Commons BY-NC.
Neil Williams is product owner for the corporate departmental publishing platform in the single domain project, on loan to the Government Digital Service from the Department for Business, Innovation and Skills.
Comment by A quick tour of Inside Government | Government Digital Service posted on
[...] Why do people need government department websites? [...]
Comment by Introducing the next phase of the GOV.UK beta | Government Digital Service posted on
[...] The INSIDE GOVERNMENT part, on the other hand, is about meeting the needs of the engaged minority. That’s fewer people (but still many millions) coming back more frequently – either because it’s part of their job or because they have a personal interest in what government is doing. They’ll be in ‘research mode’ and are likely to spend longer on the site. They also want to be notified when things change. (See our previous post about who visits government websites and why). [...]
Comment by mannybuckle posted on
Hi I just wanted to say that Ryan I totally agree with your theory about gov corporate sites working harder at being a premier resource for its news but however I would like to say I feel that the key news sites get the point across just as well.
In some instances they have the better resources and ability in which to do this. They are much more popular and people would in most instances go to this. Although the obvious place to find government news would be a government website not everyone thinks like that.
Comment by I’m not dead, I’m a dad « BASIC CRAFT posted on
[...] which includes Mike Bracken (Executive Director of Digital), James Stewart (Tech lead at GDS) and Neil Williams (who’s been quiet on his own blog to blog here). What a line up. This is where you have go [...]
Comment by Michael Saunby (@msaunby) posted on
Your three observations are totally consistent with my personal experience of managing just one tiny site of the thousands. What I honestly don't get is the single domain concept. How does this deliver benefit to anyone given that so many visitors, new and returning, come through Google searches?
I know my site http://www.metoffice.gov.uk/hadobs needs refreshing - I'm working on it. It needs more engaging content and it would almost certainly benefit from a blog and other means of supporting relationships with users - I get that. I also know that to do that I need to update the infrastructure, e.g. Drupal hosting, and some decent design (ideally by professionals) - we did this for http://www.oldweather.org and it looks good and users enjoy the site.
Comment by The vision for government corporate websites in the Single Domain (with product wireframes) | Government Digital Service posted on
[...] first defined the concept for this shared platform, then analysed users’ needs, the next step in early September was to establish and communicate the product [...]
Comment by Mark posted on
We're currently working with another Government to rationalise some sections of their corporate site.
Recent focus groups that I've run with users of the site (health professionals and patients groups) suggest that poorly structured content is a problem, but also that our concerns around purposing content for different user types would be 90% answered if only all content on the site were written in plain English.
As it is, even policy types are put off by long, difficult to read content. And generally speaking, content isn't purposed with search in mind, causing some difficulties with search.
So while structure and navigation remain a challenge for us there, the far tougher challenge is around getting content contributors and section editors to understand the real requirements for web content, and write accordingly.
Whether the programme of training that we have proposed is sufficient remains to be seen: getting content editors who are embedded within departments to feel a connection with their readers is something we aim for.
Comment by NeilF (@neilfranklin) posted on
Can't help but think that the two-way relationship that's needed is still steeped in the need for culture change in departments. It's nearly 2012 and I would have hoped we'd be further along the road to bringing other civil servants round to a more citizen-centric way of thinking than we have so far.
There's some great stuff here, but to get the best out of it, most of the heavy-lifting may need to be done away form uk.gov and with management in the departments themselves. But that could be misson creep...
Comment by Neil Williams posted on
Thanks Ryan. I think you'll like where we're heading, in that case. Definitely focusing on the "should", while trying also to keep the best of the current user experience. Stay tuned and I'll post here soon to share the starter-for-ten wireframes we've been using to aid discussions, so you can see if it's along the lines you mean.
Comment by Ryan posted on
I've always thought that analysis of stats relating to use of gov websites was too narrow in scope - limited to reactions to what it does now, without recourse to what it *should* be doing.
It's always been my opinion that gov corporate sites should work harder at being the premier resource for its news and narrative-building, rather than stay reliant on 'old' media and key news sites to deliver the message.
But this is very difficult to 'prove', as there isn't a good enough gov proposition currently going on to confirm or deny the theory. I think if this project allowed gov the facilities to provide a thorough, timely resource relating to its news and announcements, the results would be quite eye-opening.
Comment by Chris posted on
When government websites get too involved in narrative-building, there's a danger of the civil-service becoming politicized.
In a healthy democracy, the official websites should stick to facts, processes and procedures, with narrative and analysis left to politicians and media.
Before anyone says anything - that our media and politicians are exactly trusted is beside the point.