Data is everywhere and predictive analytics, using information about what has happened to predict what may happen, to influence decision making and create innovative solutions is now commonplace rather than an emerging discipline.
Most of us are aware of things like targeted discount vouchers received from our regular supermarket, where information about our past shopping habits is used to predict which other products we may be persuaded to buy. But data is being used all around us, in more innovative ways, to shape solutions provided in a growing number of sectors. Netflix, for example, uses analytics to predict the success of a new show or movie, and uses this information to decide on their schedule. ‘Predictive Policing’ helps police forces to target resource on areas that are predicted to have high levels of criminal activity. So what about Social Housing? At Orchard we wanted to know how we could use predictive analytics to deliver valuable solutions to social landlords.
Social landlords have a lot of data. My own involvement with Orchard’s Business Intelligence team has demonstrated just how much data customers have, and never more so than when we were faced with the challenge of how to load and model that data into a data warehouse in a fairly limited overnight window! So how could we use this data and what could we do?
Our initial thoughts were that we should be able to predict something, and that we should be able to use that something to develop solutions that deliver value to our customers. But what? And how?
We spoke to our customers, and had discussions internally, and started to get some ideas about the areas in which we could use predictive analytics, lots of ideas, from arrears to repairs to antisocial behaviour. So developing a platform on which we can deliver those ideas seemed like a sensible thing to do. The aim of the Analytics Platform is to enable Orchard to deliver cloud based, mobile first, analytics solutions to social landlords.
In the current environment, where organisations are looking to maximise income and reduce expenditure, predicting arrears seemed like the obvious place to start for our first solution. Currently, we are developing a solution to predict and prioritise arrears using both the rent account and external data. And how? Well, we chose to work with Newcastle University to bring in the skills and experience that we needed. We are working with Newcastle University under the government backed Knowledge Transfer Partnership. This has enabled us to develop an analytics capability within Orchard, with the oversight of academics who will help to ensure we are doing it right.
Our journey into Analytics actually started more than a year ago. Predictive Analytics was a totally new area for us, and we needed to work with someone who knew the ropes. At that stage we simply wanted to determine whether analytics was something that could be used to deliver value to landlords. We started off by engaging with Lancaster University Centre for Forecasting, working with one of their Masters Students on a 12 week project to predict repairs volumes. The aim of our project wasn’t to produce a market ready product, but to demonstrate whether predictive analytics was possible using the data that our customers have. That project was a success, we were able to predict repairs volumes, and with a extremely high degree of accuracy. Most importantly, it set us off on the path to delivering high value analytics solutions via the Analytics Platform.
I believe that analytics has a huge part to play across the board in helping organisations to tell a story from the vast quantities of data they have available. This is especially true as change becomes the norm, and ‘old truths’ may no longer apply. Our current focus is arrears, but we have a long-term vision for our Analytics Platform. with other areas (such as repairs) very much on our radar for further investigation. We look forward to helping organisations to predict their future.