Data Science is becoming more popular in the private sector, particularly in new businesses who have built their competitive advantage by being good at it. It’s time to start talking about where advanced data techniques, such as data modelling, machine learning and data storytelling can be applied in our cities to improve them and how new insights might change the way authorities approach improving the lives of their citizens.
Let me caveat the conversations we’re about to start having with our city authorities about data science by saying no one – yes, no one – in local government in the UK has mastered ‘Data Science’ right now. By this I mean, no one has recruited a team of data scientists and deployed machine learning and predictive analytics to fundamentally change how they deliver an entire service to citizens yet. Some are further ahead than others, like London Borough of Barking and Dagenham but no one has re-designed an entire service yet. Why? For a few reasons.
One: because it’s expensive to grow a team of highly skilled professionals from scratch and it’s time consuming to up-skill existing staff. Data Scientists can be masters or PhD qualified graduates; they’re statisticians, mathematicians, computer programmers and problem-solvers rolled into one. They’re skilled at turning large quantities of structured and unstructured data into implementable data-driven business strategy. They know Python, R and how to run command line direct from their console to compute data. They know how to set-up databases and harness cloud-computing to crunch hundreds and thousands of data points across exotic data-sets. More importantly, they know how to turn their domain knowledge and insights into powerful stories that can elicit policy and service change in their organisation.
The national average salary for a data scientist is £42k within the UK with high competition from the private sector.
Two: because the data maturity of local authorities is low compared to some competitive commercial businesses who have built their modern business on connecting, mining and responding to customer data in the most intelligent way possible to squeeze every last bit of value out of their customer and their operations.
Local Authorities are complex, with dozens of lines of service overlapping and interacting with the same citizens but using different software systems linked to individually recorded databases of critical service data – ranging from the meticulously recorded in modern CRM systems to free text, scanned and handwritten documents, warts and all, in those services that are stripped of cash, time, vision and the ability to convene the market to produce reliable modern products at an affordable cost.
Data warehousing programmes, or alternative initiatives to gather data together, are time consuming and costly when most cities operate in excess of 50+ IT systems across their estate and where data about one citizen can potentially exist on up to 30 separate databases with no unique identifier to connect them.
Three: because it’s difficult to describe what Data Science exactly is and the reasons for investing in it to leaders and budget holders. This is a new profession that’s forming – take a look at the trend results on the term ‘Data Science’ over the past 7-years and you’ll see it really only started taking off in 2013 and is still growing in use and intrigue — people are increasingly asking, ‘what is Data Science?’ Which might be a good thing. There are numerous descriptions across the internet of what a Data Scientist is. But when researching for this blog I could not find any stories that described it in a succinct way with examples of real benefits that have been delivered to the public sector so clear that you can say, “I want that.” Public Sector organisations in City-Regions are still coming up with their Data strategies and goals and working out they can apply these new techniques to re-design services based on analytics and predictive intervention.
So, a quick grounding in the reality of the state of where we are. But why should we not be disheartened? Because only by thinking about and expressing examples of how Data Science can be applied in City-Regions across the UK, learning from the lessons of the private sector, can we raise the topic higher up the agenda and build road-maps toward getting more advanced data capability in our City-Regions.
Here’s a list of 9 ideas to help you to start having conversations with leaders in your organisation about where Data Science could be applied in your organisation right now:
- Procurement and Commissioning — Be an intelligent customer. Use Data Science to help inform pricing and incentive payments on large outsourced contracts based on historical information and customer insight.
- Repairs and Maintenance — Estimate when you might expect to have to repair or replace costly infrastructure. Data-driven approaches to service delivery will means predictions about when things might break and the historical cost of frequent repairs vs. investing in new infrastructure — saving you money and enabling a better service to your service users. The classic example is boiler repair and replacement.
- Customer Service — Trial Virtual Assistants (chatbots) in your service. Powered by machine learning – they can improve their ability to accurately respond to customer queries over time. They can support digital inclusion and accessibility, helping those who struggle with forms to fill them in more easily. Enfield are currently exploring their use on the front line of customer enquiries.
- City Planning — Predict house price surges in areas of the city and create strategies to mitigate the effects of densifying and declining areas by comparing your internal data with private sector data, such as consumer spend, housing search results, and house price data.
- Workforce Planning — ideal for contact centres. Look at historical data to forecast likely peaks and troughs of customer activity and staffing fluctuations throughout the year. Get better at scenario planning and have good flexible working arrangements or Digital Customer Service triage in place to cover the peaks and troughs.
- Fraud Detection — Join customer data across the organisation and use machine learning to identify patterns that may indicate fraud. Highly relevant for Revenue and Benefits services.
- Interventions in social care — Make the data machine readable and securely connect it with the data held by other lines of business or even local service providers. Analyse statistics and language in cases. Compare to historical trends in order to ascertain the risk profile of certain cases. Change policy and process to ensure that a proportion of the work of a case worker switches from routine to risk-based intervention when newly recorded events trigger defined concerns.
- Managing Arrears — apply data science to better understand citizens in the context of their historical transactions. Design interventions that are more tailored to the citizen’s situation.
- Fleet Management — Understand the efficiency of your fleet of vehicles. Make sure that you procure the most fuel efficient vehicles by analysing patterns of fuel consumption depending on the type of journey’s your vehicles are most likely to be doing. Or look to purchase vehicles through companies that can provide this level of intelligence.
So there’s nine ideas to begin talking about applying Data Science to transform Service Delivery in your organisation.
If you’re already trying or thinking of trying to tackle a public sector service problem through Data Science in your city then please get in touch.