Officers from the UK’s Avon and Somerset police department are tapping into a trove of criminal records to assess a person’s vulnerability risk or offending risk. That has helped to reduce the time taken for suspect briefings, and ensure officers do not miss key details.

The police force’s use of data analytics to fight crime emerged after a few tragic incidents. Two 10-year-old schoolgirls were murdered in 2002 by a man who worked as school caretaker even though the police had records of him being a sex offender. In 2013, an Iranian refugee was beaten to death and set on fire despite multiple calls for help to the police.

The force has since learnt from those incidents and ramped up its analytics efforts to solve challenging policing problems with the help of Qlik, says Charlie Farah, Senior Strategic Client Advisor of Qlik. He discusses four crucial steps for predictive policymaking.

1. Decide where to focus efforts

Many agencies often need to “take a step back” and be strategic about which areas they want to focus on, Farah says. “Rather than trying to address every little issue with predictive policymaking, identifying which areas will give the biggest impact” will maximise resources.

Some governments have established data frameworks to collectively determine where to place their efforts, he adds. That’s generally the “best approach”, as that stops governments from investing in programmes that do not provide significant benefits.

Fortunately, technology enables organisations to test, explore, and fail fast, he says. A data integration and analytics platform can help governments test hypotheses quickly, determine the impact of projects within a matter of weeks, and decide whether or not to continue with them.

2. Break down collaboration silos

Today, many silos within agencies have been broken down, Farah says. Most government departments have employed Chief Data Officers to encourage data sharing within the organisation.

But the current challenge is around the sharing of data between agencies, he adds. “What we’ve seen with Covid-19 is that the quicker we can get access to data, the more effective our decision making is going to be.” Integrating multiple data sets from different agencies will give a complete picture of how challenges can be tackled.

Some agencies don’t have the capabilities to cope with the volume and velocity of data as well as where to find the data, Farah says. Qlik has a “data cataloging system” that provides users with access to analytics-ready data – helping them get insights and answers accurately and quickly.

It also helps organisations integrate and prepare raw data for analysis, he adds. That allows them to “make the best decision possible without having any delays.”

A school in the US is working with Qlik to identify students at risk, Farah shares. Qlik’s platform integrates multiple datasets from the education, social, and police departments. If a particular student has low attendance rates, the school can find out if there have been any calls from the police department to the family residence to determine if there are hints of domestic violence or other matters impacting school attendance and performance.

3. Tap on human expertise

Predictive policymaking is not about “having the technology spit out some really high-end insights and then expect everyone to take it as gospel,” Farah emphasises.

“Predictive models can help people generate the initial insights, but then humans, using their experience and intuition, can validate those insights and interrogate them to make the best decisions.” People should not be taken out of the equation, he adds.

Qlik has a “cognitive engine” that gives data engineers a peripheral view of data , allowing users to freely explore their data without the limited “tunnel vision” Farah says. That allows them to have a holistic view of data and from their experience and intuition, decide on the best course of action.

4. Create a data-driven culture

Results from Qlik’s recent survey revealed that 68 per cent of public sector organisations in APAC needed more data literate staff, Farah says. Organisations need to arm the workforce with the right opportunities and tools to enhance data interpretation skills, he adds.

That doesn’t only require the right technology, but also a culture and mindset shift, he emphasises.

A top-down message from the executives will make staff “feel comfortable about challenging decisions with data”. If, for instance, data reveals that a traditional approach is not the most efficient, having the right culture will make it easier for staff to suggest alterations.

Predictive analytics can help to close the gap and ensure governments deliver efficient and responsive services. With the right collaborations, human expertise, and culture, this might just be the future of government.