From Post-Mortem to Prevention: Rethinking Data Innovation in Public Protection

Last week, I attended the National Police Chiefs’ Council (NPCC) Innovation and Digital Summit 2025— in the exhibitor space common to all such gatherings was much talk of transformation, efficiency, and the promise of technology to make policing and public services smarter. The atmosphere was optimistic, with plenty of stalls offering the next big thing in “data-driven insight” or “digital collaboration.” However, as I moved from conversation to conversation, a troubling pattern emerged.

Almost every solution on display was built around reacting to harm, not preventing it. They were impressive, yes—tools capable of visualising crime hotspots, linking offender histories, streamlining investigations and collating digital forensics. But their focus was squarely on the aftermath: after the crime, after the crisis, after the damage was done.

This is the paradox of innovation in public protection today. We are rich in systems that tell us what went wrong but have very few that help us understand what might go wrong before it does.

The people I met were sincere and committed, and their solutions technically brilliant in many cases. They showed dashboards that glowed with real-time updates, algorithms that could simplify trends, and reporting tools that looked like something out of a sci-fi film. Yet underneath the digital polish, the data was the same familiar kind—incident reports, crime logs, case files. Information triggered and collected only once something bad had happened.

That’s not innovation; that’s post-mortem analysis.

The reason for this isn’t laziness or lack of imagination. It’s the structural and technical silos that still exist between agencies. Police, justice, health, education, social care, housing—they all hold crucial pieces of the same human puzzle. But each piece sits in a separate system, bound by its own data standards, consent models, and governance rules. Add in leadership priorities and accountability that might not dovetail comfortably, and fragmentation is inevitable.

This problem runs deep across public services. Schools often know when a child’s attendance is slipping long before police or health services are aware of emerging risks. Housing officers might spot patterns of anti-social behaviour or rent arrears that hint at domestic instability. Health professionals see the early symptoms of stress, neglect, or substance misuse. Even mental health and physical health are treated by different services with poor information sharing in some cases. Each agency or service sees only their narrow slice of the picture. Without a mechanism to bring those fragments together, we are routinely left with disconnected interventions that arrive too late to change outcomes.

Somewhere in somebody’s system, we know everything about people after they’ve entered the criminal justice system or after they’ve reached crisis point, and we are trying to do our best for people at this point. However, across all those systems, we probably also had the jigsaw pieces that could have signalled risk in time to make a difference…if only someone was putting the puzzle together.

This is what I kept coming back to in my conversations yesterday. The waste is not just financial—though the cost of reacting rather than preventing is enormous—it’s human. It’s the lost potential of children who could have been supported earlier, families who could have been stabilised, communities that could have been safer. Every reactive dataset is a missed opportunity.

That’s why my time at the summit with Capita gave me some hope. I should stress that, whilst I have no vested interest in their product and was not paid to be there, I was there at their invitation. I was there to share my experience of using a forerunner of their multi-agency data sharing (MADS) capability to design and commission a violence reduction intervention in Milton Keynes – an intervention that had a measurable impact on reducing school exclusion for the participants we were able to target using effective data sharing across three organisations.  

Capita were talking not about simple data dashboards or analytics in isolation, but about an approach to MADS that allows organisations to bring their data together securely, ethically, and intelligently. Instead of waiting for a crime report or crisis referral, genuine data sharing within a coherent partnership enables those partners—local authorities, police, NHS trusts, schools and others—to pool relevant data sets and analyse them collectively, identifying risks, patterns and correlations before they harden into harm. Crucially, partners can then use these data to collaborate on doing something much smarter than we usually do.

It’s not surveillance or reporting; it’s intelligent prevention.

Having seen their presentation and read their literature, I do believe that MADS could be a genuine game changer. However, whether agencies or organisations choose to use Capita’s product or not, the potential value and impact of its principles are undeniable. Frankly, it is either a collective shame or – to be more generous – a systemic failing that those principles are not already at the core of all our work to improve public health.  

Clearly, what makes this approach powerful is the ability to overlay multiple risk indicators from multiple partners and develop a shared understanding that comes from knowing what others previously kept in their ‘silo’. Take a teenager who has increasing school absences, a family known to social services, and a recent mental health referral. Individually, each of those signals are, sadly, all too common to really raise much alarm in the current context of our services. But seen together—through shared data—they tell a clearer story of vulnerability and potential escalation.

That’s the holy grail of early intervention: not waiting for a safeguarding referral or a crime to occur but spotting the constellation of small signs that predict bigger problems.

Of course, most people working on the front line in these services know this. However, at present, complex ethical and legal challenges around sharing data, alongside concerns around consent, proportionality and privacy tend to cause paralysis; too many leaders remain risk-averse to the point of inaction when it comes to sharing data with other agencies. However, not sharing has consequences too and it certainly doesn’t tend to make people safer. Instead, it usually makes them invisible until they’re in crisis.

This is why the shift toward proactive, preventative data use is so urgent. The technology exists. The appetite to do something differently exists. What’s been missing is the framework that allows agencies to work together in real time, guided by shared principles of purpose, transparency, and accountability. An effective MADS model does this and offers the promise of the sort of network of harm prevention that can help us make better decisions earlier.

At the summit, there was a lot of talk about efficiency. Every organisation wants to do more with less, to automate, to streamline. But the greatest inefficiency of all is duplication: the endless cycle of separate agencies reacting to the same family, the same crisis, the same harm, but from different angles, with different data.

Early intervention—real early intervention—requires a different mindset. It means building partnerships rooted in trust and shared accountability and it means employing technology that doesn’t just connect systems but connects missions. We need to be braver about shifting investment and attention to the front end of the process—to those early, quiet signals that something is beginning to go wrong. That’s where the potential for real impact lies.

Digital innovation cannot always be about making our reaction times faster or taking the humans out of the process; it should also focus on reducing the number of things we have to react to in the first place.

And that’s only possible when data stops being a series of post-crisis snapshots and becomes a live, shared, preventative ecosystem.