Putting Data to Work: 7 Questions with London’s Eddie Copeland


I recently sat down with Eddie Copeland, Director of Government Innovation at Nesta, a U.K.-based “think tank and do tank” whose mission “is to bring great ideas to life.”

At Nesta, Eddie is a change-agent; he is collaborating with Andrew Collinge, Assistant Director at the Greater London Authority, to run pilots for a London Office of Analytics, similar to the work I did with Mayor Bloomberg in New York City.

As part of my Putting Data to Work blog series, I look forward to sharing thoughts from people getting things done. People who are putting analytics and data to work. People who break through obstacles to make things happen. On that note, here are some outtakes from my conversation with Eddie.

Mike: What is the goal of the London Data project?

Eddie: We were inspired by your work with the Mayor’s Office of Data Analytics in New York City, and now we’re working with the Greater London Authority to test whether that model can be adapted for the very different context of London. The aim of this process is to prove to both central government and city leaders that if local government and public sector bodies are willing to share, analyse and act upon their combined data, with the support of an expert team of data scientists, then they can achieve things they couldn’t do if they were working alone.

1606_LODA_CopyrightDavidAltabev2016_FINAL_LR_6Mike: How did you determine where to start?

Eddie: Your advice was clear. There are lots of pressing urgent public sector challenges that face every city. But if you’re just starting out on a data analytics journey, you have a far better chance for success if you start with something very specific. It may not be as glamorous, but by starting simple it can help build momentum for more ambitious initiatives further down the line.

Mike: Yes, that was a real cautionary note I needed to impart. It can take years and years to tackle issues but if right from start, you don’t know what you’re trying to accomplish, you won’t go anywhere. This isn’t to say the problem is “solved” with the first project; indeed it rarely ever is solved in year 1 or year 5. The goal is always to provide the organization better vigilance and situational awareness. What can paralyze organizations – government and private sector alike – is endless deliberation over whether a proposed solution simply pushes the problem elsewhere rather than addresses its root cause. The intelligent use of data renders this a false choice – as you dive into a discrete part of the problem, the ideal solution is one that will enable both a greater understanding of the problem and the ability to follow it as it moves around in response to your efforts to address it.

1606_LODA_CopyrightDavidAltabev2016_FINAL_LR_3Eddie: We took that advice on board. We started by crowdsourcing from the boroughs a list of potential public service challenge areas that could be tackled with data. Andrew Collinge then led a workshop to test them against the criteria for running a successful data pilot. The result of that was a decision to focus on tackling landlords of HMOs – houses of multiple occupancy – who fail to license their properties correctly. HMOs are properties rented out by at least three people who are not from one ‘household’ (e.g. a family) but who share facilities such as the bathroom and kitchen.

Mike: How is data going to help you?

Eddie: Data in aggregate is more powerful than siloed data. By joining resources and using data from each borough, we believe we can build a predictive model that can help focus inspectors’ time on the right buildings. Similar to your work in New York City, that data can help us determine all the possible things that may correlate with high-risk HMOs. There’s also the potential to include data sets from City Hall, or the Fire and Rescue service. Instead of training the model on just one borough, we’ll have a far richer set of data to work from.

Mike: Now that you’re a few months in, what are the biggest lessons learned to date on this project?

Eddie: The biggest lesson we’ve taken from our conversations with you is to relentlessly focus on actionable insights. A tendency of much of the smart city movement is to put the emphasis on the technology and data. Instead, we know we have to focus on the outcomes we want to enable. You had a common sense message.

1606_LODA_CopyrightDavidAltabev2016_FINAL_LR_2Mike: Question one is: What do you want to do more effectively or differently? The only way to win people over and succeed is to have a real, actionable plan in mind.

Question two is: What information product do you need? Is it a map that shows you where to go; a widget that combines different data sets; or a prioritized list of places to inspect?

If you can’t answer those two key questions, then maybe data analytics isn’t the right solution for your problem. Did you face any reluctance from the boroughs to participate?

Eddie: No. We’ve been delighted by the level of enthusiasm to participate. UK local authorities and the wider public sector are facing significant financial challenges right now, and need to find ways to do more and better with less. I think the majority now realize that some form of data-driven collaboration can play a role in that.

The key point is that we need to help them on that journey. Typically, public sector bodies are  held back from collaborating because of technical barriers, data silos, organizational silos, and real and perceived legal barriers. Overcoming the latter two are the hardest. The reality is you can do a lot of things with data with a change in psyche and culture, and if you do it securely with the right permissions.

Mike: What will success look like?

Eddie: By the end of 2016, it will be that we’ve run a successful data pilot that proves that if you join multiple data sets, you can come up with an insight – and act on that insight – to improve a public service. The next step is likely to be looking at how we can expand the pilots – either to more boroughs or to cover a greater number of issues. The dream by the end of 2017 would be to work with Andrew Collinge to help deliver a permanent London Office of Data Analytics (LODA) in City Hall – though its final shape and function will depend on what we learn through this fascinating process.

To read more about Eddie’s data analytics journey with the City of London, visit http://data.london.gov.uk/blog/

Photo copyrights to David Altabev