The transactional services list we published two weeks ago was the first time Government had attempted to gather data on all of its services together. Clifford Sheppard, Data Analyst, explains why categories were chosen to manage the data and some of the stories hidden inside.
When we began the task of gathering this data together, we did not know exactly how much data we would have, or how difficult it would be to help people make sense of it. As the list grew ever longer, we decided that it would be helpful to group the services according to their main function.
As well as helping us to manage an unwieldy list, we wanted to look across Departments for common themes and patterns. We also wanted to make sure that GDS were supporting a good range of projects. Categorisation helped us answer questions like: What was the primary purpose of most Government services? Were high-volume services more likely to be a certain type than low-volume services? Did some Departments specialise in particular service types?
Taking a sample of services, we defined the main function of each service using nine simple categories, including requesting benefits or grants, booking appointments, requesting permissions or licences, and reporting information.
We found out that:
- about 30% of services were requesting permission or a specific licence to do something
- another 30% involved providing or reporting information to the government - these services accounted for over two-thirds of all completed transactions
- only 1% of services - but over 10% of transactions - involve making payments to Government
We were surprised at some of our findings. Instinctively, we had assumed that payments to Government would account for quite a lot of services, but that group covered a relatively small number of high-volume transactions.
We’ll be doing more work with categories before our next release. For example, we’re looking at categorising high-volume services by the technical components they use, such as logging in, processing payments, and postcode lookups.
What do you think? Have you done something similar with your services? We’d love to hear your ideas on how we can categorise transactional services and find more of the stories hidden in our data.