High IT spends do not always result in the best performance, as data from 163 associations shows. John Wickenden crunches the numbers
Information technology and communications is an essential outlay for any business, and an activity that can vary considerably in cost.
Conventional anecdotal wisdom about ITC relates the cost variation to geographical spread and operating structure, though this is often not backed by reliable evidence.
As part of its annual cost and performance benchmarking exercise, Housemark has been collecting optional detailed overhead measures on central services such as finance, human resources and ITC.
ITC measures include the number of offices served, the geographical spread of offices, server uptime and user satisfaction. All of these measures can be cross-referenced with staffing levels and costs.
Using 2010/11 data submitted by 163 English housing associations, we conducted a statistical correlation analysis of cost and performance to test how some common assumptions about ITC fared:
Servicing far-flung offices leads to higher ITC costs:
In spite of a moderate positive correlation between hardware support costs and the ‘furthest presence’, distance made no real difference to overall ITC costs.
Lots of offices are expensive to support:
There was nothing more than a weak correlation between the number of offices served and the cost of ITC per property, the cost per user, the cost of hardware support or the size of the ITC team.
Larger organisations invest more:
Using the number of ‘properties managed’ as a proxy, there was no consistent correlation with investment in hardware, software or total costs. The size of the organisation made no difference to the relative amount spent on ITC.
User satisfaction comes at a price:
Our data showed no correlation between user satisfaction and the total ITC cost per property, the cost per user or the average cost of a PC.
100 per cent server uptime is difficult to achieve on a budget:
Our data showed no correlation between the amount of time servers are operating for and ITC total cost per property, the proportion of ITC employees or the cost of support.
The data and analysis suggest that these five factors have little impact on the cost of running an ITC service. So what causes the variation in costs?
In-built obsolescence is the norm in technology and as a result, hardware purchase and depreciation is a particularly large outlay for social landlords. It can account for up to 5 per cent of landlords’ overheads - and annual costs vary considerably depending on reporting policy and asset replacement cycles.
Another driver is application and software support; this forms the largest proportion of total ITC costs and is highly variable - from nothing to more than £3,500 per user. Supporting users is often the most difficult part of an IT service to provide as, like insurance, paying for too much support offers poor value for money, whereas too little cover can end up costing more in the long run if something goes wrong.
The analysis suggests conventional wisdom is flawed. Spend on ITC does not correlate with performance, and geography does not directly influence cost. The key drivers seem to be procurement and policy - both largely under associations’ control.
John Wickenden is knowledge manager at Housemark
housing associations in England that took part in the survey
median total IT spend per user in 2010/11
minimum total IT spend per user in 2010/11
maximum total IT spend per user in 2010/11