In January a board paper put our technology spend at a little over four percent of revenue, against a sector benchmark of just under three. The paper recommended a review of IT costs, and the review was mine to lead, so I began by finding out what the benchmark actually measured.
The survey behind it asks companies to report their information technology budget, and leaves each of them to decide what that is. Ours has accumulated a great deal over fifteen years. It holds the telephony and every mobile contract in the company, including the drivers' handsets. It holds vehicle telematics, the dispatch printers and their labels, and the licences for the warehouse management system. It holds about ninety software subscriptions bought by other departments, which landed on our cost centre because we were the only team with a purchasing card for software. The three comparable companies I could get to talk to me put most of those things under operations, fleet or the department that uses them.
Restated on their basis, our spend came to slightly under three percent, level with the benchmark. Nothing had been cut. We had only moved lines between two definitions.
I would not have found this if the number had looked good. Nobody had questioned it for years because it had never been above the benchmark, and I had never checked whether the comparison was fair until it stopped being flattering.
What goes to the board now is a figure in three parts: the core technology any company would recognise as IT, the technology that runs our depots and fleet, and change spending on projects. Each part carries a short note of what is included, and the second part is shown per order dispatched rather than as a share of revenue, because it grows with volume rather than with sales. Departmental subscriptions are recharged to the departments that chose them, which reduced their number by about a fifth within six months without anybody being asked to cancel anything.
The review did find savings, about ninety thousand pounds a year, mostly in the subscriptions. It found them because we were looking at what the money bought rather than at the ratio.
A benchmark compares two definitions before it compares two companies. Ours had been comparing a warehouse's technology with a head office's.
– Serguey Shinder
Top comments (0)