Introduction

This paper presents examples of ‘data visualisation’ using data from different Housing Associations across the UK. Each visualisation played a part in helping the Registered Provider (RP) concerned to reduce its costs. The purpose behind the paper is to demonstrate some of the insights about value in organisations that are possible using data from within the sector. It will hopefully encourage RPs to explore their own data in more depth and to create similar insights as a stimulus for change.

1. Keeping track of arrears

Figure 1 below shows for each type of tenancy agreement the distribution of customers with improving arrears, and the distribution of those with worsening arrears. Each line is a different tenancy agreement and each dot is a customer in arrears. If a customer is to the left of the centre line their arrears are improving, and if they are to the right their arrears are worsening. The closer they are to the edge, the faster the rate of change. This chart is from a dynamic model which has revealed, for example, that the most rapid rises in arrears have taken place within certain types of starter agreement. Tenancy type is therefore a key feature within predictive analytics aimed at identifying the potential for arrears growth early on.

2. Field Officers driving less

Figure 2 below represents two days in the life of repairs operatives. Each dot is a visit to a customer’s home. Each line is a journey for one repairs team. The thickness of a line represents the distance driven. This provided the starting point for a simulation of the benefit that better scheduling methods would bring. When the simulation was put into motion, the number of longer journeys diminished and the productivity of the majority of teams was shown to increase.

3. Helping customers keep their homes

In Figure 3 below, each circle is a customer. The dark circles are customers who have been evicted or abandoned their home. The light circles are current customers. The strands indicate common features that link them together. The size of the circles indicate the size of any outstanding arrears. The closer the current customers are to the former the more likely they are to follow suit and also suffer eviction or abandonment. Those customers presenting the greatest risk of eviction or abandonment are now flagged for immediate investigation. Whenever either outcome is avoided significant cost savings are realised.

4. Contact with the customer is vital

Figure 4 below indicates that arrears can be improved as long as contact with the customer is maintained. Customers with significant arrears were clustered according to the strength of contact maintained with them. Each line represents a cluster and its distance up the chart reflects the strength of contact with those in the cluster over a 3 month period. Those at the top of the chart were regularly in direct contact. Those at the bottom were not, either because they chose not to respond, or because contact effort could not be sustained. If a line shows an upward trend, the average arrears balance for the cluster improved. Although this produced few surprises (the weaker the contact, the harder it is to recover arrears), it confirmed that predictive analytics to spot future arrears cases should include predictions for levels of contact as a key independent variable.

5. Finding the optimum component costs

Figure 5 below shows an increase in the average repairs per kitchen per year the longer it is before they are replaced. Although total repairs increase once new kitchens are installed, the average number of repairs per kitchen is lower than for later years. In other words, many repairs but across many properties in the early years. Fewer repairs but also across fewer properties in the later years. The average repairs per property reporting repairs therefore increases. The boundary lines either side of the trend line represent the 2 & 3 standard deviation limits respectively. Although significant data quality issues had to be overcome e.g. the age of older kitchens could only be determined from Stock Condition Survey data, this analysis permitted an optimum replacement age for kitchens to be calculated. This is the age at which the total cost for repairs and replacement is at a minimum.

6. Freeing up time in Finance

Figure 6 below demonstrates levels of activity within part of the Finance Team at month-end. Each circle is an activity performed by the Management Accounting Team at month-end. The thickness of the lines indicates the amount of interaction that takes place between the activities. This analysis triggered time-logging by the team during every month-end for 6 months. Its purpose was to find ways to shorten the time taken and to free time for other things.

Figure 7 shows the results. Each pie represents the logging data for one of the 6 months. Each segment is one of the activities in the chart above. Its size reflects the amount of time captured for the activity. The red segments are activities regarded as delivering little value. Efforts to shorten the time required for month-end focused upon these first of all. The green indicates how much time was freed as the time needed for the red activities steadily reduced.

If you would like to know more

We at Develin have helped a large number of organisations across many different sectors to manage their costs down through the use of highly effective analytics. If we can help through direct support, training, or a combination of the two, please don’t hesitate to get in touch.