This morning, I was stuck by a story on BBC Radio Four that highlights the strength of benchmarking and demonstrates real insight from segmenting data.

The story was about a study conducted by the MOD and published this morning in the Lancet. Service personnel ‘more likely’ to commit violence after a tour. The headline statistic is that veterans commit violent crime (11%) more than the benchmark of the national average (6.7%).

The MOD admitted in the piece the widely held belief that once trained to be comfortable in harm’s way soldiers stay comfortable in harm’s way. Unfortunately, soldiers are better trained at harm than the general population.

The real insight comes from segmenting the data to indicate where the focus of resources should be applied. Looking at the data for younger ex-soldiers, under 30, violent offending is a stark 20.6% of offences, compared to the 6.7% average. A point missed by mainstream media was that this group were only 2728 men. These men have been trained and subsequently left the army early, for whatever reason they have not developed into career soldiers.

Having left the army, I fear that these young men are falling between two very different worlds. Rather than talk about what the statistics say about ex-servicemen, we should be talking about 2728 individuals, a number that society does have the resources to approach directly and build a focused action plan.