Visualising data in local government
A conversation about the use of data visualisation in local government operations.
Data visualisation tools such as Tableau, PowerBI and Google Data Studio are slowly infiltrating local government. So what insight can tools like this produce?
To better my understanding I ask Ben Proctor, technical director at The Satori Lab, what he thinks.
Luke: What types of data do local governments collect that could be visualised?
Ben: It might genuinely be easier to identify data that they collect that couldn’t be visualised. But let’s think of some broad areas of data:
- Data about how services are working: how many people apply for a thing, how many are successful, how long does it take for an application to be processed, how satisfied are they by the process, how much difference does it make to their lives and so on
- Data about people: what is the demographic breakdown of the council area, of a neighbourhood, or a village, what needs do different people have, how rich or poor are they, how fit and healthy and so on
- Data about things: where are our dog poo bins, where are the roads, the schools, the community centres, the parks and so on and then you can relate those different types of data to each other… how well do our services work for different sorts of people, or people in different places, does the presence of a dog poo bin affect dog mess problems and so on and then you can add in time, has our service got better or worse for this type of person over the past year, does this bin fill faster than that bin, how is the population of children changing over time near that school
And that’s really only the basic stuff…
Luke: OK, so we can clearly see that there is data available, but how do local governments ensure it is good quality data?
Ben: Imperfectly.
There is a vicious/virtual circle with regards to data quality (not just in local government but certainly there).
Poor quality data cannot be relied on by the service, so the service is managed in other ways, so they don’t collect good quality data.
Good quality data can be relied on by service, if they use the data to manage the service this incentivises them to improve the data quality.
Many councils don’t use data to manage their service so they are stuck in the poor data quality zone. (It’s not just me saying that, the WAO recently said that sort of thing about local government in Wales).
Luke: Which teams will typically visualise the data?
Ben: Like all matters pertaining to local government there is no simple answer and every council is different. Place based services (planning, bin collections, highways etc) tend to be comfortable with maps. If they are based in a large council or are lucky they may have a GIS (Geographic Information System) specialist available to help them. If not they’ll have to muddle through with whatever their Management Information System allows them to visualise on maps.
Accountants live in spreadsheets and they tend to be very comfortable spilling out financial charts. These days many managers also have some familiarity with spreadsheets and so they will usually be reasonably free with some simple charts.
Across other services there will be a much more mixed picture depending on the priorities of the profession, the skills of the individual and the prevailing culture of the council.
In larger local authorities there may be a dedicated Performance / Business Intelligence team. In principle the people in this team will be more skilled in visualising different sorts of data. They *should* be able to bring together different datasets from different teams to create deeper insights. Often, however, they are hampered by the quality of the data from individual service areas and the lack of appreciation amongst managers of the value that can be generated by bringing different datasets together, and then analysing and visualising them to make them useful to services.
Luke: So what things do councils need to do to get better at visualising their data?
Ben: I think we need to take a step back here. Visualising data shouldn’t be a goal in itself. The goal should be improving services, data can help everyone improve services, visualising data can help make it more useful.
Often the conversation jumps straight to the visualisation and therefore to the tech (and there’s lots of lovely tech). The starting point should be the service though. My recommendation for all teams and organisations wanting to start getting better at using data is to start with people on the frontline and ask them a few simple questions:
- how do you know what is the most important thing for you to do next?
- how do you know that you’re doing a good job?
- how do you know whether your team is doing a better job than last month?
The first question tends to point you in the direction of being able to visualise the whole service, so the sorts of visualisation that work well there are simple charts showing demand building up for different team members. This helps team members work out whether they should pick up another new job from the queue or going and help one of their colleagues clear the backlog. This is the sort of thing that Kanban can really help with in some environments.
The second question relates to the design of the service. What need is the service meeting and how effectively is it meeting that need. Ideally when services are designed the design should include some meaningful measures that can help to answer this question. This sort of thing is ideal for a dashboard.
The third question relates to constant improvement. Ideally the team needs to come together regularly to review what went well and what they need to do differently. This sort of thing requires more detailed analysis and for most councils is the time to bring out the spreadsheets (and involve your Business Intelligence team if you are lucky enough to have one).
I work at The Satori Lab.
