“One necessary qualification for anyone in management is to stop asking people to explain the ups and downs (day to day, month to month, and year to year) that come from random variation.”
Dr Brian Joiner
The importance of data and especially its security has come to prominence of late, not just in the housing sector but also in high-profile cases in the media. The arrival of GDPR this month is also something to occupy all responsible organisations who handle personal information. But what about how we use data in understanding and improving business performance?
This is not related so much to personal information but to performance data – the kinds of data that populate senior management dashboards – and crucially the efficacy of this data is its use as alluded to by the quote above. This kind of data usage should have more prominence – it is the difference between running our businesses using the best available information on which to make decisions or poor, sometimes meaningless, data or in some cases little to no data at all. In order to be effective, efficient and to provide value for customers, housing providers need to have data at the centre of their improvement plans; indeed, improvement must be data driven.
This is actually using data to make the business better, reducing your bottom line and actually improving customer satisfaction. McKinsey’s Global Institute indicates that data-driven organisations are 23 times more likely to acquire customers, six times as likely to retain those customers, and 19 times as likely to be profitable as a result. It will take an investment in time, energy and capital, but being data driven will provide excellent returns on the investment.
So what is it to be data driven? Effective business decisions rest on good data, well used. Without data, decisions can only be based on guesses, gut feelings, opinions or something even less scientific. By ‘good data’, we mean data that is effectively collected, using appropriate methodologies, and properly governed. Data must be reliable in order to enable good decision making – that is the first step. But the main premise here is that it is what you do with the data to help you learn from it that is of paramount importance. Good data improperly handled and analysed is as bad as poor data.
First, data is only as good as its method of analysis. It was Chris Argyris who conceived of the ‘ladder of inference’. Argyris recognised that the data we select from what we observe is influenced by our beliefs and assumptions. We therefore need a method of analysis that helps us to focus on and gain a knowledge and understanding of the right things in order to achieve our purpose. Data-driven improvement is first and foremost based on the underlying principle that the key to continuous improvement is learning, gained by studying service systems and performance data in the right way. Change should be based on evidence and the quality and use of the data on which this evidence is based is crucial to success.
Many organisations use data to create performance indicators presented in dashboards. The way data is presented is a major influence on any subsequent decisions. The example of the common use of binary comparisons is a case in point. This is something explored by Simon Guilfoyle in his book “Intelligent Policing”, a forensic analysis of poor use of data within the Police Service. What Guilfoyle explained is that deriving any meaning from the comparison of two data points, which may be monthly or annual comparisons of averages, is unlikely to be useful in terms of making informed decisions.
For example, if we are measuring something that results in a number that we don’t want to rise (such as void turnaround time, rent arrears or time taken to complete repairs), taking action based on a binary comparison would be foolish. If the number had increased, let’s say, from 19 to 35 between months or years this tells us little about the pattern or variation in performance over a period of time.
The use of time-series data creates visibility of performance and its variation over time to enable a deeper understanding. Any decision made based on a monthly or annual binary comparison is likely to be unsuccessful at best, and possibly damaging.
So to conclude, to be truly data driven, improvement must be based on the right measures. Many housing providers rely too heavily on data derived from poor measures, or measures that have limited ability to provide useful information. According to John Seddon, creator of the Vanguard Method for Service Improvement, the test of a good measure (i.e. its effectiveness) is that it:
- Relates to purpose;
- Demonstrates variation over time;
- Facilitates learning and understanding (i.e. leading to knowledge about the system);
- Is used by those doing the work to monitor, control and improve the work;
- Is used by managers to act on the system (i.e. predict, plan and make best use of resources).
Many of the measures in use within housing wouldn’t pass this test. There are still many measures being used based around targets. For example in housing repairs, completion times for repairs are often measured against target, generating data which tells us the percentage of repairs completed within and out of target for urgent, routine and emergency repairs. This is very limited because it doesn’t directly relate to purpose from a customer’s point of view, nor does it demonstrate variations over time, nor facilitate understanding.
Having the right IT systems for data collection and storage and indeed good quality data are all important, but unless there is the know-how within the organisation to make good use of the data, they are less likely to translate into useful improvement actions and more likely to become meaningless numbers.
Tim Brooks is a senior consultant at Data Futurists.
