On Friday 19th December 2014 when the final agency switched on its pages we celebrated GOV.UK being ‘organisation complete’.
Sat in a chippy dressed in a dinner jacket with heroes of the @GOVUK transition. Supremely proud to be part of this thing.
— Benjamin Welby (@bmwelby) December 19, 2014
Three years ago one of the four things Baroness Lane Fox told government to do was ‘fix publishing‘. She recognised that hundreds of different publishing platforms could do a good job in isolation but required the public to understand the complexity of government and that approaching similar needs in bespoke ways was expensive and inefficient. It wasn’t the first time government had recognised the complexity of its web estate and we’ve stood on those broad shoulders to successfully replace over 600 websites with just the one.
That achievement is only really the end of the beginning but I’ve been reflecting on my highlights so far, in anticipation of what’s to come. I’ve got seven.
1. The summer of 2012.
The transition project took 15 months but GDS’ work on the single government domain started with DirectGov and Business Link.
I joined GOV.UK in May 2012 after first Alphagov and then GOV.UK’s beta had launched and spent a lot of the summer poring over Business Link. Every one of its £105m idiosyncrasies made me thankful that the government was building its own team that wouldn’t be restricted to the terms of an outsourced contract, n days of development activity and limited releases of code each year. The reality (of course) is that GOV.UK has ended up with flaws of our own but we are dealing with them, and it didn’t feel like that was true about Business Link.
We could easily don rose tinted spectacles and consider that Olympic summer our halcyon days: we were a smaller team with a smaller product operating with the freedom of being disruptive new kids on the block. But we were also chancing our arm – that we’re still here now is down to proving the point.
It’s also down to those of our colleagues who spent their time and energy shielding us from people in suits and cajoling, challenging and charming them to think differently about digital services and entrust the nation’s publishing platform to people in t-shirts.
The trust that got built over that summer, and launching GOV.UK without incident that autumn, is the foundation for everything GDS has been up to since.
2. Closing feedback loops.
These days GOV.UK is pretty big – consistently seeing over 9,000,000 unique visitors a week. We spend a lot of time working to better meet their needs but sometimes our users need additional help.
When that happens there’s a team who respond to their queries to make sure that people get the support they need.
Before the transition of the 312 agencies and arms length bodies started in earnest I remember worrying how this small team might cope with a doubling of traffic and the increase in the diversity of the audience. I didn’t need to: the user support team for the ‘organisation complete’ single domain is about the same size as it was then.
That’s been possible because as more agencies have arrived our handling of feedback has got more sophisticated. There have been spikes in the contact rate but every time someone contacts us it’s an opportunity to isolate problems and solve issues so they don’t affect the next person with the same need. And we know that’s working because, for every million visitors to GOV.UK, only 360 of them need extra help, seek clarification or report a problem.
It’s a small, perfectly formed and quite, quite brilliant team. They’re a very unsung but absolutely crucial part of the last two years.
3. Protecting URLs.
New websites mean new URLs and the old ones can get forgotten, littering the internet with broken links. We’ve put a lot of effort into avoiding that. Those first months at GDS poring over Business Link were an early part of that. I was crafting an almighty spreadsheet to ensure we had mappings for every old URL. All 77,000 of them.
My spreadsheet fed our earliest redirection software. As with everything else on GOV.UK that process and our software has been continuously improved. Today there are no spreadsheets and each organisation manages its share of the 1,800,000 mappings using our Transition tooling. That’s amongst the most important (and hidden) things underpinning GOV.UK and ensures anyone who comes across a link to one of 1,500 domains will get to the correct content, or a signpost to the record held by the National Archive as appropriate.
— Benjamin Welby (@bmwelby) December 14, 2014
Over the last few weeks we’ve been thinking about how GOV.UK should handle content before, during and after the election. We’d hoped to share some of the detail before Christmas but you can’t rush something as sensitive as how GOV.UK reflects changes to the governance of the country, especially now we’re the only place government publishes.
It still a highlight though because it’s something made possible by the single government domain. The research we’ve done and the prototypes we’ve built show that the economy of scope and scale from one website instead of hundreds can help us do the right thing for our users, and government content.
It’s also been a brilliant endorsement of the iterative way we work. We’ve sat with users in front of the current site and understood how it’s not meeting their needs; we’ve made changes, shown them to other people and repeated. Our latest thinking has been warmly received so we’re hopeful it’s going to ease the pain for one particular group of our users.
5. Starting with user needs.
The centrality of design principle number 1 was clear when I first walked through the door at Aviation House. But whilst it runs through our veins we have struggled to maintain the relationship between a need and the associated content once it’s published.
We used the Need-o-tron (RIP) to record citizen needs, business needs were managed in a series of spreadsheets and then the departmental transition identified generic user needs that could be met by different content formats.
Importing thousands of pieces of content into those formats based on a point in time didn’t always embed user needs in those organisations. With 312 agencies bringing their content to GOV.UK we knew their hundreds of editors could become advocates for a user needs led approach to designing content.
So we ran a lot of training – every agency went through an introduction to user needs, and then every agency participated in a workshop where they took analytics data, call centre insight and other user research they’d done to identify and record their users’ needs.
I sat on the floor in the corner observing our first attempt at that workshop. It didn’t go well. We went back to the drawing board and tried again until we we found a formula that worked. One of my favourite days at GDS was the one we spent in Caterham at the home of an agency editor trying out that new approach. I don’t know how many workshops we ran in total (one colleague alone led an epic 55 of them) but they ensured each and every agency transitioned to GOV.UK having put their user needs first.
Now that GOV.UK is ‘organisation complete’ our database of user needs, Maslow, should contain an almost complete hierarchy of needs met by government. It’s got 3,237 and the recent launch of /info/ pages is one way that we’re opening up, and re-using that information to make GOV.UK better.
You can’t really overstate the importance of taxation to the fabric of society and HMRC’s green website was one of the most familiar outposts of the government. That site has 620m pageviews a year and 120,000 URLs covering over 40 different taxes and benefits.
You couldn’t ship that overnight and so, over the last year and a bit a team has been chiselling it away piece by piece and incrementally redirecting things from the old, to the new.
As you might imagine it’s involved close working between different people at HMRC and different people at GDS. But it hasn’t just been a GDS content designer’s challenge (although a lot of words have been written by them). There have been lots of words re-written and re-purposed by people inside HMRC. Different teams within GDS have built new GOV.UK features without which transition would have been impossible and HMRC have made use of existing GOV.UK functionality to improve the experience of their users.
The vast majority of HMRC’s site (in content terms) is found in the tax manuals. These redacted versions of the documents used by tax inspectors to do their job have been made available under the Freedom of Information Act and are an essential piece of information for tax professionals. Again there’s been good teamwork – GOV.UK teams have built a new manuals frontend and an API while a team at HMRC has been building the backend that HMRC editors will use to keep the manuals up to date.
I spent a few months on the HMRC exemplars and saw first hand some of the challenges of legacy contracts, legacy technology and legacy attitudes. Those challenges are largely absent for GOV.UK: we’ve been a team built from scratch, building a thing from scratch. It’s easy for us to forget but I think GOV.UK has got better at being not just part of GDS but part of the government team. On the day hmrc.gov.uk became gov.uk/hmrc, GOV.UK and HMRC stood shoulder to shoulder, as a single team using a single government domain.
— Mike Bracken (@MTBracken) December 16, 2014
7. All the people. So many people.
I’ve recently had a job interview to spend some more time on GOV.UK, the outcome of that’s unknown but whether I’m part of its future or not I’ll forever be proud to have been part of this and be in awe of the women and men who got us here.
I’ve not worked closely with all of them but you couldn’t ask for a more passionate, committed and intelligent group of people to be building and improving something that is, and will continue to be, for everybody.
So Merry Christmas, happy new year and a massive thank you from me to this lot (particularly whoever’s supporting the site while we’re enjoying ourselves over the next 10 days!): Aaron K, Abigail W, Adam M, Ade A, Alan M, Albert M, Aleks M, Alex J, Alex M, Alex T, Alex T, Ali B, Alice B, Alice N, Alice W, Alyson F, Amy K, Amy W, Amy W, Andrew L, Andy H, Andy K, Angela H, Anna L, Anna P-S, Anna S, Anne C, Ashley S, Ashraf C, Ben A, Ben C, Ben H, Ben T, Benedict S, Benjamin P, Bob K, Bob W, Brad W, Brij T, Brooke S, Caley S, Camille B, Carl M, Caroline J, Cath R, Cathy T, Charles D, Charlotte C, Charlotte D, Charlotte H, Cheryl T, Chris H, Chris M, Chris M, Christine C, Clare B, Clare L, Clare M, Clare Y, Cliff S, Dafydd V, Dan C, Daniel C, Daniel R, Darren H, Dave F, Dave M, David H, David I, David L, David S, David T, David W, David W, Davie T, Davina S, Dipa S, Dom B, Donna F, Ed H, Edd S, Elena FdR, Eliot C-M, Elisse J, Emer C, Emily D, Emily W, Esteban O, Etienne P, Felicity S, Fiona R, Fola E, Fran L, Fran W, Frances B, Francis M, Gareth R, Gavan C, Gavin C, Gaynor B, Giles T, Glynn M, Graham F, Guy M, Gwen C, Helen N, Henry H, Hinrich von H, Holly G, Ian H, Ian, Isabell D, Jack B, Jack S, Jaimes N, Jake B, James A, James C, James L, James S, James S, James T, James W, Jamie A, Jamie C, Jane E, Jane O’, Janet H, Jeffrey P, Jenni M, Jenny D, Jeremy H, Jo L, Joe L, John B, John O, John T, Jon S, Jordan H, Josh R, Julian M, Karen F, Karl E, Karyn R, Kathy S, Katrina D, Kaye S, Keith E, Kelly A, Kieron H, Kieron, Kiran P, Kushal P, Lana G, Leigh M, Leisa R, Liam O’, Lil B, Linda E-M, Lindsey K, Lisa S, Liz G, Liz H, Liz J, Liz L, Lou K, Louise S, Lucy H, Lupe N-F, Lynne R, Léonie W, M Norm F, Maria D, Mark H, Mark H, Mark O’, Mark S, Mark S, Martha L-F, Martyn I, Mat W, Matt B, Matt F, Matt H, Matt S, Mazz M, Mel P, Meri W, Mike B, Michael B, Mike H-R, Monica K, Natalie S, Natasha B, Nayeema C, Neil M, Neil W, Nettie W, Nick B, Nick C, Nick S, Nick S, Nikki A, Nikki M, Ola A, Oliver S, Olivia N, Padma G, Paul C, Paul C, Paul D, Paul H, Paul H, Paula H, Paula S, Pauline F, Persis H, Pete H, Peter H, Peter I, Peter J, Peter U, Phil T, Philip P, Poss A, Ralph C, Ray K, Rebecca C, Rhiannon B, Richard B, Richard P, Richard S, Richard T, Ricky M, Rob N, Rob Y, Robin C, Robin M, Roo R, Rosalie M, Rosie C, Ross F, Ross W, Russell D, Russell G, Russell T, Sam R, Sam S, Sara B, Sara V, Sarah P, Sarah R, Sarah W, Sheila B, Simon K, Spencer B, Stephanie D, Stephen E, Stephen G, Stephen K, Stephen M, Stephen R, Stephen W, Steve L, Stuart C, Subhi B, Sue D, Sue K, Tara L, Tara S, Tekin S, Thayer P, Till W, Tim P, Tom A, Tom B, Tom B, Tom H, Tom L, Tom S, Tommy P, Tony S, Toria L, Vicky B, Vinay P, Wendy C, Will C, Will R, Will R, Yvonne C and all those I’ve forgotten or whose contributions I don’t know.