I'm puzzled by the way that some people in the ICT sector still tout benchmarking as the most effective way to adjust continuing services contracts to deliver value for money. In my experience, the best way to achieve ongoing value for money (VFM) is to adjust little and often as opposed to conducting major infrequent re-baselining exercises.

Companies spend a disproportionate amount of time thinking about VFM at the outset of a project, while ensuring that VFM remains over the duration of the relationship gets less attention. A little time devoted to this issue pre-contract can pay huge dividends.

There's no single magic bullet mechanism to ensure that a services relationship maintains its overall competitiveness in the market. Companies need to construct a series of mechanisms whose combined effect will be to maintain (and maybe improve) VFM. Too often, companies simply rely on the existence of a benchmarking clause and then simply hope for the best.

But major, infrequent benchmarking exercises are much less effective than small, regular adjustments at re-aligning the contract to improve VFM. It's much better to build these adjustments into the daily life of the outsourcing relationship.

Start by using the governance process to create the atmosphere in which VFM can be improved. From that process should flow some clearly identifiable mechanisms that are tuned to the nature and scope of the outsourcing.

Good VFM is a blend of techniques. These include ensuring the provider is obliged to make continuous improvements to implemented processes; incentivising the provider to make cost savings; using a gainshare model to incentivise improvements in technology and processes; and implementing productivity measures or gradually improving productivity baselines.

Benchmarking does have a role as a contract tool but many benchmarking clauses are written in hope rather than reality. Benchmarking tends to work much better for more commoditised services where there is an easily comparable market. More complex services that are blended across service towers in different countries can be hard to benchmark.

Companies should be realistic about benchmarking and only use it when it stands a chance of working well or as an input to a regular, pre-planned discussion about service parameters. And don't expect benchmarking to be easy, argument-free, or lead to a straightforward, no-hassle way of getting price reductions.