Data is a currency: spend it wisely.
Chris McLaughlin, managing director, MIS AMS
Data is becoming a currency, but how do we get value from it? We leave trails of data wherever we go these days – each time we click on a website, or buy something, sign up to a newsletter or interact with an organisation, we produce data that, when aggregated and analysed can reveal trends about our behaviour. Used in the right way, providers can then deliver us personalised services we actually want – which of course, grows their profits. Win-win.
It’s a challenge the housing sector is coming to terms with tackling. Collecting data is one thing, using it to best effect in decision making within organisations is an entirely different challenge. As a whole, the sector is slightly behind the curve in terms of collecting and using data to predict trends and thus to deliver tailored housing association services for tenants.
So what can housing providers learn from building up data over time? What trends can they identify and how might that information help them to provide the ultimate housing solution for tenants? In 2015 a group of housing associations put their heads together and decided to pool their data in order to understand more wholly the trends within the sector. Small housing providers may only have around 1000-1500 assets that give a micro view of the market place as a whole. The housing charity HACT and a group of 16 housing providers with around 400,000 homes between them collaborated so that the volume of data would provide insights. Putting their heads together meant learning much more.
Rent arrears & repairs
When pooled in a vault in the cloud with strict security protocols so that residents data couldn’t be identified, the providers were able to cross reference rent arrears information about tenants with other sets of data such as geographical location, types of property and information about the tenants themselves in terms of how often they access services from their provider. This created a way to spot underlying issues that when highlighted early enough, could be addressed to reduce the number of tenants falling into arrears. Prevention is always better than cure. Predicting that tenants fitting a specific profile might default on their rent payments gives the provider an ability to stop it happening before it becomes a problem. It highlights those tenants that need the most support before they actually need it, and importantly for the provider, gives them a starting point that may help reduce rent arrears.
In the same way, predicting repairs in certain types of properties is finally becoming a reality. As we collect more data, it’s possible to determine in what type of property to invest in the future based upon return on investment and longevity of use. Some housing associations do a full stock review looking at what their costs are, consolidating any loans and validating repairs against the asset. If the rent collected is £50 per week and repairs cost more than £200 each month, housing associations are more able to decide what type of properties bring best return by accounting for the depreciation of their assets, and investing more wisely in the right type of properties going forward. For bigger associations it provides efficiency savings, but it’s not beyond the pale to suggest that for smaller associations analytics are increasingly important – especially those in deprived areas where there is very little room for manoeuvre in rent versus costs. Even down to the basics – when will the boiler need replacing, the house need a new roof, or the windows need updating – it’s all costs that stack up against the rent that’s collected, and ultimately leaves the association with difficult decisions to make. Predicting costs more accurately thanks to data analytics helps manage this before the costs are incurred and allows a change in strategy to counteract them.
Of course with every up-side, there comes the challenges. One might suggest that if we are increasingly able to predict rent arrears or even anti-social behaviour from tenants more easily with trends in the data we analyse, then there is a greater risk of penalising or even denying a tenant their tenancy. It’s easy to argue that the increased information allows us to better support the tenant, but there is always a risk it may be used in the wrong way. Could landlords rent properties to one tenant over another based upon their likelihood of some of these wider issues? It throws up some uncomfortable questions. Likewise, if a tenant is currently unemployed and spends more time in their property, is there a connection with an increase in repairs due to more wear and tear on the property? The data may suggest so, providing the landlord with the evidence to choose tenants based upon the information to hand. In the main, I believe that landlords will responsibly put their estates in order, rather than penalise their tenants. Understanding the difficulties simply means housing associations are better equipped to provide the right advice and course of action before its critical.
Cost benefit analysis
Of course, the costs of big data must be weighed up against the benefits of mining the information for decision-making purposes. Storing data in large warehouses is often an expensive affair and can become more costly if not done correctly from the beginning. The use of a data warehouse allows several products to access data from several systems without having all the information in one single database. But a data warehouse that hasn’t been assembled correctly, can be a big headache. In housing it’s often used for reporting or for passing data been CRM, asset management and/or mobile devices. For integrated suppliers like MIS AMS, the point at which a customer tells us they have a data warehouse is often a pivotal moment with potential additional risks. If systems show differing values that simply don’t marry up, costs can spiral.
There are many components to consider: there is the amount of storage space, maintaining the information, consultancy fees, hidden costs in terms of time, the mobile costs to achieve it and ultimately the data must be put to good use. And of course, our biggest fear is always the cost and time it can take to run relevant reports from the data warehouse. There must be a serious amount of computational power to run reports freely from that amount of data backed up over years. I wouldn’t recommend running it over the main company network as that would quickly render it useless for hours – hence the associated costs of an entirely separate infrastructure just for data analysis. The list of costs can be endless, and are they worth the output?
Future Internet of Things
As we link more devices to the Internet and they generate increasing amounts of data, it’s not difficult to predict that in the future big data and analytics will become more central to business strategy. But to eradicate costs, sifting of data will be increasingly important. Which parts of your data do you want and need to extract, versus the parts that are less valuable to your business? Which data will provide the most return on investment? Where do you draw the line? Like any currency, it’s important to make sure that it’s valuable to you. My advice to housing associations is look at the big picture and focus on where changes could be made to bring about efficiencies, and gradually introduce these rather than doing too much at one time.