Service Manual

When will more people visit GOV.UK using a mobile or tablet than a PC?

Yesterday the BBC published data showing more people accessing iPlayer via tablet than via computer. This prompted me to update some of the data I gathered for the government’s agreed approach to mobile last this time last year.

The objective of the UK government’s digital strategy is to make sure our digital services that are so straightforward and convenient that all those who can use them will choose to do so. But what our users consider to be ‘straightforward’ and ‘convenient’ is not static. We need government services to be able to adapt quickly to big changes in people’s behaviours and expectations.

For example, here’s a graph showing how the devices people use to visit GOV.UK have changed since its launch. (To be precise, the data is for visits, rather than users.)

Percentages of visits to GOV.UK from computer, mobile and tablet
Percentages of visits to GOV.UK from computer, mobile and tablet


Since 1 January 2014,  63% of visits to GOV.UK have come from a computer, 23% from a mobile and 14% from a tablet. In January 2012 it was 77% computer, 15% mobile and 9% tablet. If you visit the GOV.UK performance dashboard you’ll see that the sample sizes are non-trivial.

Compared with the general UK population, the graph above may be skewed by a minority (around  2%) of GOV.UK users who visit the site more than 100 times a month, often to research government activity as part of their job, typically from a work computer.

I’ve tried to get more representative UK data by looking at the visit data for the two weeks following Christmas Day, when such power users are probably not quite so busy.

The device breakdown for this period last year was 74% computer, 16% mobile and 10% tablet.

This year saw 61% using a computer, 24% mobile and 15% tablet.

On Christmas Day 2013, only 51% visited GOV.UK from a computer, compared with 66% on Christmas Day 2012. (Over 300k visits to GOV.UK this past Christmas Day; 34k were looking for a job; over 5k bought a tax disc.)

Shifts in the devices people use to access the internet should come as no surprise, but the pace of change might. And I do not expect this switch away from PCs towards more personal, portable, touchscreen devices to slow down anytime soon.

The UK government e-petitions service has seen incredible changes in how, when and where it is used. Pete Herlihy has product managed this service since it went live in summer 2011. As he revealed recently, only two years ago over 75% of visits came via computers. Now a mere 27% do so, with 56% from mobile, and 17% from tablet.

Not every service will end up with such proportions, but e-petitions demonstrates just how rapidly and radically user behaviour can change. Here’s current data for some of the transactional services on GOV.UK:

Book your practical driving test:

Computer – 67.4% (was 71.3% in March 2013)
Mobile – 21.4% (was 17.7% in March 2013)
Tablet – 11.2% (was 11% in March 2013)

Change date of practical driving test booking:

Computer – 56.9% (was 61.3% in March 2013)
Mobile – 32.4% (was 30.3% in March 2013)
Tablet – 8.7% (was 8.4% in March 2013)

Apply for a Student Finance:

Computer – 64.6%
Mobile – 26.6%
Tablet – 8.8%

Make a Lasting Power of Attorney:

Computer – 84.3%
Mobile – 12.4%
Tablet – 3.3%

Apply for Carer’s Allowance:

Computer – 67.1%
Mobile – 17.7%
Tablet – 15.2%

I hope this helps explain why the digital by default service standard requires that, from April 2014 onwards, all new or redesigned central government digital services must be designed with an appropriate range of devices in mind. As we say in the GDS Design Principles, our services must understand the context in which people will use them. And for many people, for many services, that context is swiftly becoming more mobile, more personal and more touch-controlled.

Designing for small screens can be a real challenge. Which is why for many of the 25 exemplar services we’re now designing the mobile version first, despite visits from computers still being in the majority. Why? Simply, it’s often easier to make a service also work for a computer monitor and keyboard if you’ve already made it work really well on a small touchscreen than it is to go the other way.

Moreover, as Andy Washington, MD of Expedia UK & Ireland, explained at a recent panel, designing within the constraints of a small touchscreen helps keep your underlying service as clear and as simple as it needs to be to serve all your users, including those who may be new to the internet, or find it a struggle.

Finally, to answer the question posed in the title to this post: When will more people visit GOV.UK using a mobile or tablet than a PC? On Christmas Day 2014, if not before.

NB I’ve seen no data over the past year to suggest the government’s approach to downloadable apps should change. We’re still not ‘appy about them, and central government departments and agencies must seek an exemption before they start developing any.

How many people are missing out on JavaScript enhancement?

A few weeks back, we were chatting about the architecture of the Individual Electoral Registration web service.  We started discussing the pros and cons of an approach that would provide a significantly different interaction for any people not running JavaScript.

“What proportion of people is that?” an inquisitive mind asked.


We didn’t really have any idea how many people are experiencing UK government web services without the enhancement of JavaScript. That’s a bad thing for a team that is evangelical about data driven design, so I thought we should find out.

The answer is:

1.1% of people aren’t getting JavaScript enhancements (1 in 93).

That’s not the whole story though. Read more.

Learning from assessments

We’ve learnt a lot about assessing services over the last few months. James Kemp, acting Service Manager for passport appointment booking at HM Passport Office, attended one of the early assessments and we asked him to write about the experience.

As one of the very first services to go through the assessment process it wasn’t entirely clear what I needed to provide in the way of evidence, or how strongly we would be held to the published criteria.

So, in the run-up to the panel, I read through the Government Service Design Manual (which I strongly recommend) and the 26 service standards making notes against each one on how well I thought we met them and what would demonstrate that. I thought that we managed to meet the vast majority or had action in hand to meet them within a week or two. There were about five that I thought we might struggle with; I hoped that would be sufficient for the service assessment.


Talent scouting

Today we published another important update to the Government Service Design Manual to help departments recruit and retain staff with new digital skills. This is one of the services the new GDS Recruitment Hub is offering departments.

Increasingly we need to bring in people with specialist skills that the Civil Service hasn’t been able to call upon before – designers, web ops, product managers, and the like. We’ve created template job descriptions for a set of leadership and specialist roles and we’ve provided organisational design guidance about how people in these roles should work together. We’ve also given advice on salaries. (more…)

Giving the manual some style

Over the last couple of weeks keen followers of the service manual will have noticed a few changes to some of the guidance.

Along with publishing pieces like the assisted digital action plan and the guidance on increasing digital take-up we’ve started revisiting everything we published since the beta and bringing it up to style.

Many authors, one voice

Back when we first started the alpha of the manual we expected to publish every piece of guidance on the site with the name and contact details of its author. It quickly became clear that this wasn’t going to work, as guidance:

  • often had many authors
  • needed to be able to be frequently revised
  • had to stand alone, without one-to-one support
  • needed to feel like something anyone working on services could contribute to

Besides which, the Github repo does a fairly good job of capturing that information for those who really wanted to know the details. We already link to that on every page of guidance in the manual.

Getting up to Standard

The Government Digital Strategy says that all new or redesigned digital services will meet the Digital by Default Service Standard from April 2014. Before then, GDS is using the standard to support the Cabinet Office spending controls, to deepen knowledge of the standard across government, and to identify areas where we can support departments in building their digital capability.

We’ve now started to carry out assessments against the standard. I want to explain how and why we’re doing these, and how you can request an assessment for your service.

The 26 articles of the Digital by Default Service Standard were published alongside the Government Service Design Manual a few weeks ago. Formal assessments of new and redesigned services against the standard will begin from April 2014, however, it’s important we start using the standard as early as possible. Partly because it gives the teams building services time to get up to speed with the requirements, and also because it will give us feedback that we can use to fine-tune the standard over the course of the year, and improve the guidance we offer through the manual.