Peter Jordan blogged recently about the new content dashboards. This week John Joseph from the Department for Environment, Food & Rural Affairs (Defra), writes about how he used Google Analytics and a little help from GDS to create dashboards to track and improve statistics at Defra.
At Defra, we need good analytics data to understand what people want from our statistics, and to build a more agile service. I created a dashboard that was easy to develop and costs nothing to implement. I thought others might be interested in doing the same.
This was not what I originally set out to do. After we moved to GOV.UK we needed basic analytics reporting on the Defra statistical content. That was my original goal. The statisticians wanted monthly updates, tracking over 200 publication pages. Doing these updates manually using Google Analytics proved time consuming, cumbersome and soul destroying. I don’t have a background in analytics and no prior knowledge of common programming languages….I still don’t!
John Byrne at GDS said the Magic script add-on in Google documents might alleviate some of my suffering. He was right. I followed a short video tutorial that helped me through every stage of development. GDS got me started, but within a couple of hours I was getting the hang of it. This was great as we had limited time to devote to this project and no money.
I created automated reports quickly, but the video described moving on to creating a shared dashboard. I thought I could deliver something better than we originally set out to achieve.
We are pretty satisfied with our statistics dashboards now. It’s not perfect and there are a large caveats in the data (mainly around sampling levels), but it was quick, cheap, agile and effective. We can now provide real-time critical information back to all our statisticians.
The dashboard requires no maintenance as the updates are driven automatically within the add-on. And it cost us nothing, apart from some time to develop.
It is also providing some real insight we can use. For instance, our analytics show a significant number of our visitors are infrequent users who are not exactly sure what they are looking for. They tend to enter a generic phrase like ‘waste statistics’ which takes them to a navigation page. We are working on the signposting, content and navigation to help users find what they need faster.
The dashboard also provides some rapid feedback on the issues that are trending. We are working out how we can find these spikes of interest and fix user journeys by linking to other topical information.
The approach offers huge potential. We are so pleased with the results that we are starting to roll this out more generally. We aim to devote more time to rapid analysis and interpretation of all this data we are now collecting to build better and more responsive content focused around customer needs. I hope this tale inspires others to do likewise.