British Army enhances data-driven decision making to staunch churn

The British Army has been improving the quality of its personnel data in recent years, in part with a view to reducing officer churn.

“We have helped the Army understand why it lost some [high value] personnel to voluntary outflow – soldiers deciding to leave before the end of their contracts – and how it could work to prevent this,” said SAS, the analytics supplier the army has been working with on its data since 2011, in a statement.

Tim Carmichael, chief data officer and chief analytics officer at the British Army, said: “The success of the armed forces will always be the result of close collaboration between soldiers, policy-makers, civil servants and suppliers such as SAS.

“The personal judgement of our most experienced staff will continue to reserve the final say in every decision we make. Yet, in whatever task we bring data to bear, we have seen greater successes achieved. SAS has helped us improve the quality of our decisions – not replacing, but augmenting the human judgement at the heart of the Army.”

As reported in Computer Weekly in 2011, the Army’s data quality improvement programme began in earnest in December 2008, with a data management and governance programme headed up by Detica, a consultancy firm owned by BAE Systems. One upshot of this was a contract with SAS for a full set of data mining and reporting capabilities.

The data problem that the Army Personnel Data Management Organisation was then confronting was not knowing exactly how many soldiers they had, and at what stage they were at in their cycle through the army.

By 2013, the Army’s data management programme was facing a fresh challenge, with a "whole army" concept – unifying regular and territorial army soldiers into one force – announced by Philip Hammond, then secretary of state for defence, in November 2012.

SAS, at that time, claimed the army had saved £770m by using software, including SAS Enterprise BI Server, a business intelligence software system that integrates SAS Analytics and SAS Data Management, and had done so by analysing and more efficiently allocating manpower and other resources.

The supplier has now said that “using a large data set of every soldier to voluntarily leave the Army within the last five years, we determined that the most likely motivations a soldier would have for voluntary exit would be their age and length of service, often coinciding with a change in the size of their family”.

It continued: “The model has proven instrumental in helping staff officers identify the conditions that could lead to the early exit of valuable personnel, allowing them to take pre-emptive action to encourage the soldier to stay.”

Since initial deployment, adoption of its platform has, the supplier said, expanded to 700 users in the army today. “While primarily used by planners and policy makers, SAS also sees significant use by logistics, education and investment teams as well as for sentiment analysis of the workforce,” it said. 

The army is using SAS Visual Analytics and is now using SAS Operations Research to help it optimise processes and personnel deployment. It has also recently approved a proof of concept for SAS Text Analytics, which it hopes will allow it to use open source data and more efficiently process freedom of information requests and paperwork.