Crime doesn’t repeat itself by accident, it repeats because the same people keep causing it…

Early in my career, I remember standing in a store looking at footage of yet another theft.

It was all very familiar.

You reviewed the CCTV, logged the incident, circulated an image, and moved on to the next problem.

Then a week later, the same individual came back and did it again.

And again.

Nothing we were doing was technically wrong. But it also wasn’t really working.

That was probably the first time it properly clicked for me that we weren’t managing crime. We were managing incidents. And I still think that’s one of the biggest issues in the sector now.

We’ve improved the tools over the years. We’ve digitised systems, improved coverage, made reporting quicker, and built better ways of gathering evidence. But in a lot of environments, the core model is still the same.

An incident happens, it gets recorded, evidence is gathered, the paperwork gets done, and then everyone resets until the next one comes along.

That creates plenty of activity. But activity is not the same as control.

Because if all we ever ask is “what happened?”, we miss the far more important question, which is “who keeps causing this?”

That’s where the real problem sits.

Across retail and public space environments, the same pattern comes up again and again. A relatively small number of offenders are responsible for a disproportionate amount of the harm.

Most people working in the sector know that. They’ve seen the same faces, the same names, the same behaviours. But very few operating models are genuinely built around that reality.

And that’s where I think we still need to shift our thinking.

If the same people are returning over and over again, then the question is not how we get quicker at reacting to each individual event. The question is why we aren’t doing more to understand, track, and manage those individuals in a structured way.

That means moving away from pure volume reporting and towards pattern recognition.

It means linking activity across time and location instead of treating every offence as if it lives in its own little box.

It means deploying resources with purpose, not just presence.

And it means thinking more like an intelligence function than a response function.

At a high level, the direction of travel is obvious. The more effective models are the ones that build a clearer picture of repeat and prolific offenders, connect incidents properly, prioritise effort against those causing the greatest harm, and work closely with policing and partnerships to drive actual outcomes.

That doesn’t mean throwing everything out and starting again. It just means making what we already do more intelligent.

And that matters now more than ever.

Crime is more persistent. In many cases it’s more organised. Police capacity is stretched. Customers are under pressure to demonstrate that they are in control. And yet too many people are still investing in models that look busy without fundamentally addressing the source of the problem.

That’s why you still see such a big gap between security spend and security outcomes.

For me, there’s a real opportunity here for the sector to move on a bit. To stop obsessing over incident counts, activity measures, and hours on site, and focus instead on reducing harm, managing risk, and delivering meaningful outcomes.

Because I still think back to that store from years ago. Not because what happened there was unusual. Quite the opposite. It was completely normal.

And that’s exactly the issue.

We do not reduce crime by getting better at recording the same incident ten times over. We reduce crime by understanding and managing the people behind it. Everything else should come after that.