Martin Hansen, director of information and network services at Nottinghamshire Police, has a warning for business intelligence audiences: "Everyone at BI conferences talks about capturing customers, but I don't think that you really want to be one of our customers…"
While the focus of business intelligence and reporting systems may be different to those in commercial organisations, the benefits that police forces expect from BI are not dissimilar.
One, says Hansen, is faster, more accurate reporting: BI systems need to help forces to deliver aggregate crime figures to the Home Office each month. As that process improved, analysts were now spending less time processing numbers and more time producing analysis in the form of 'National Intelligence Model' outputs.
As in business, analysis depends on the quality of the data. "Counting crime is a lot more difficult than you might think, as the nature of it can change from the time when it is first reported," says Hansen. "The difference between criminal damage and attempted burglary can be quite a fine line and an incident initially reported as a crime may be re-classified on further investigation."
Information is delivered to senior members of the local force more regularly. As well as actual crime rates, they increasingly want to assess how well they are fairing against key performance indicators. While crime figures are updated every day at Nottinghamshire, detection rates are reported monthly to give a wider picture of successful investigations.
"Daily reporting of crime was the key thing – the [senior officers] like to see how any particular crime is building up or dipping." Officers can drill down from overall force level to Divisional, Local Command and Beat levels to gauge the impact of their crime-fighting initiatives. "It has enabled us to understand a lot better and a lot more quickly what is happening," says Hansen.
One scheme, which banned violent football fans from every pub around a stadium if they had been banned from one, was particularly successful – and with faster reporting police able to see the impact within two weeks. In another case, an increase in minor house break-ins was linked to an increase in car theft as thieves would simply force front doors and steal car keys left on hall tables. Because these two crimes are classified separately, the trend might previously have gone unnoticed.
Such information is fed into a business intelligence dashboard (built using technology from Business Objects) which displays the number of crimes and plots these on maps green, amber and red to indicate a reduction or increase in crime. Having identified trouble spots, police officers can then start analysing the mass of unstructured data (police logs, officer reports, statements, and so on) to determine the best cause of action. Using the information gathered by the National Intelligence Model, officers can also profile problems, markets and potential targets at a country-wide level.
The detail can even be focused down, for example, to individual pub car parks where satellite navigation systems are being stolen regularly from cars, particularly between 8pm and 9pm when their owners are eating dinner. "There's no point prowling around at the wrong time of day," says Hansen. "Criminals are often creatures of habit – and it is often their undoing."
And BI is now key to linking that habit to 'customer capture'.