ONS Digital

Hello there. I manage the Publishing, Content Design and Digital Content teams at the ONS in Digital Publishing. Each week I attempt to give an overview of my work, key projects we’re involved in and how we’re trying to improve the ONS website and its products. You can find me on Twitter.

You know you’re lucky to have a great team when you’re presented back with a piece of work you commissioned and the result is far better than you ever hoped for. That was the case for me this week when I saw the results of a review of Publishing processes carried out by one of our publishing officers. Over the last few months my colleague Jordan Young has been revisiting the goals which underpinned our centralisation of publishing in the ONS in 2015.

It marked a huge shift in the process of publications at the ONS and has had some big implications – it underpinned the creation of a publishing team, led to the emergence of our Content Design team, required a whole new CMS to produce the content, and required a whole new workflow for the organisation to get their outputs from raw data into live publications.

I often feel the publishing process, with all its complexities, is mis-understood and under-supported in the organisation. It takes a huge effort and enormous co-ordination to publish the 462 bulletins, 290 articles, 153 methodology articles, 1,252 data set pages that went live in 2019. And not forgetting the hundreds of adhoc data sets pages, Freedom of Information (FOI) pages, working papers, product page changes, corporate content, and much more besides.

Jordan’s review has clearly established – thankfully – that centralising has proved both more efficient, and is considered to have improved quality, versus the previous model.  It’s also revealed that the tools we use to manage publishing, the content management system, the ways we share and engage across teams are all in need of improvement. There’s lots to do but we have so much more clarity thanks to the work.

Elsewhere this week, I spent some time working with a performance analyst to provide some analytics to colleagues looking at how well our outputs are working for users. I love spending time in Google Analytics, and it’s easy to get lost in the numbers, but together we’ve been able to provide some senior colleagues a view of performance across hundreds of our outputs.

One thing that definitely jumped out at me, was the high percentage of users downloading one of our data sets who do so without ever visiting or using one of our statistical bulletins. Given we know that 75% of bulletins are 3x to 10x longer to read than most people spend on the page there are legitimate questions around value we should be asking.

We’ve been able to split the consumption of our outputs by audience type by using a proxy – IP address. Our previous performance analyst set up a series of audience groups by looking at IP ranges. So we have home broadband, business broadband, government, local authority, police, schools and colleges, universities etc.

It’s been helpful to land the message that so many of our users  are ordinary citizens. In fact 75% of visits to the ONS website are from users on a home broadband connection or mobile connection.

Rear view mirror

More Census discussions this week: Census is looming large in the rear view mirror and those of us interested in making sure our products out of Census are as good as they can be are exploring how best we can work together. We know we need resource, we have great ambitions, and there’s just the small matter of co-ordinating teams and agreeing roles and responsibilities.

This week proved somewhat frustrating when I had to put on hold to bring in a developer to work on a data navigator tool to help users source data by theme when often datasets are spread across multiple government websites. We had a brief, an outline work plan and a group of keen stakeholders only to discover a very similar product has just been released by one of the teams at ONS.

One of the fundamental motivations in creating the Digital Publishing division was to end the practice of adhoc digital development and to ensure what is built is made to the right standards through the right process.

Clearly governance isn’t working as well as it could; so rather than starting work on a new product, we’re instead doing an audit of development happening across the organisation to better understand needs and motivations, because this isn’t the first time this has happened.

Right now, as the week ends, I feel a little like Sisyphus watching the rock roll down the hill. I’m sure, come Monday, I’ll be eager to start pushing again.