How we’re working to transform the way we work with data
Making sure we get the basics right — and looking to new ways of working
I recently joined Citizens Advice as Head of Data Science. It’s a new role, and its creation reflects the scale and ambition of what we’re trying to achieve: to transform the way the organisation works with data.
We’re starting from a position of strength. No one else sees as many people with so many different kinds of problems. This means our data gives us a unique insight into the challenges people face in England and Wales today, and we use this to improve services and inform our advocacy work. It also means data is taken seriously here. Our data analysis is valued by a variety of audiences and gives us a unique voice in the policy world.
So, our work is all about the opportunity to make things better. We’ll work to amplify and support the good work people are already doing. We’ll introduce ways of working that are new or not widespread within the organisation. We’ll improve how we work with data so we can be an even stronger voice as we speak up for our clients.
This blog is a starting point for the journey we’re embarking on at Citizens Advice, setting out our plans so far.
What’s data science and why’s it important at Citizens Advice?
It’s exciting to be working somewhere where there are potential applications of data science techniques for the public good. Data science is a varied field that uses scientific methods, processes, algorithms and systems to extract knowledge from data.
Citizens Advice generates large volumes of unstructured data — data that isn’t organised in a pre-defined way, usually with lots of text — through our work for clients, while the advice content on our public website produces a high volume of interactions with people across a wide variety of often complex topics. In both cases, we’re working with more material, more data than we can manage by hand. In 2018/19, for example, our public website was visited more than 29 million times, and we helped 1.3 million people face-to-face.
To tackle this data we can turn to techniques such as natural language processing — getting computers to analyse and process human language — to help us make more out of what we already have. An example of this in use could be in being able to identify trends in issues our clients are experiencing more quickly through automatically analysing what they’re telling us. By identifying these trends, we’d be able to better support our advisers and improve our services for clients.
To use data science to meet the size of this challenge we’re going to need new skills, new ways of working, and new tools and technologies. We’ll need to further establish data science as a discipline that works in multidisciplinary product and service teams — the teams developing and running the services we provide. We’ll also need to take a more multidisciplinary approach to building data products so we can make the most of our range of expertise and develop products to best meet the needs of the people using them.
From experience, I know it takes time to build capability like this. It’ll be difficult because the skills we need are in high demand and, as a charity, our resources are more limited. However, the mission of this organisation is strong, and it’s a supportive and motivating place to work. We don’t need to rush, and we can learn from the experience of others. We also have lots of good people already, so we’re not starting from scratch.
Tackling our data architecture
Citizens Advice doesn’t have an established practice of data architecture. This means we have limited understanding of our data structures and the various systems they sit in. This situation isn’t unique to Citizens Advice — it’s a common problem, but not understanding your data leads to duplication of effort and difficulties in joining things together. The result of this is that it can be disproportionately difficult to deliver seamless services to our users — both clients and staff.
This isn’t something you can fix quickly. Data architecture needs to be a collaborative part of developing and running services, not an arms-length exercise. We’ll progressively raise our level of understanding of the data we work with, and build or buy systems in a way that considers the wider organisation. We’ll design our data for re-use, and encourage that re-use internally and externally by working with data users.
Improving our data culture
Starting with a revitalised data strategy, we’ll develop better ways of talking about data. We’ll be more specific in our language, and more mature and informed in our decision making. We’ll support staff to better understand their responsibilities for the data that they work with, including relevant ethical considerations.
Being patient and deliberate
Citizens Advice is large and very busy. Our team is small, and the work is broad — data has a legitimate stake in every aspect of an organisation’s work. You can’t do everything at once.
We’re helped by Citizens Advice having a clear strategy to help us prioritise our efforts, and we’ve started a programme of practical improvement that’ll help us to show examples of how we can work better with our data.
I hope I’ve set out the scale of this exciting challenge. I’ll be sharing our progress regularly — but feel free to get in touch if you’d like to know more.
