I’ve always approached business with a fairly simple rule: look at the facts first. I like data. I like evidence. And whenever we’re trying to improve something at Evolve, the first question I ask is “What does the evidence actually tell us?” so when we decided to take a proper look at our recruitment practices, we started with the stats.

Research by the UK Gov showed that nearly half of firms operating in and around the cyber sector said they struggled with skills gaps. But what really caught my attention was that 33% of hires were reportedly made through “word of mouth”. Now, that’s not necessarily a negative in my mind, networking is a valuable tool to create links and share information and organisations like Women Empowering Defence are a brilliant example of how building the right connections can actively help under-represented groups find opportunity.

The issue, though, is when word-of-mouth becomes the dominant or only route into a business. If you rely solely on who’s already in your orbit, you inevitably end up hiring people who look and sound a lot like the people you’ve already got. Not deliberately, but structurally. So networks are good and they’re healthy, but they need to run alongside something more fair and more intentional.

That’s when we began tightening up our recruitment design. One of the biggest changes was around salary banding. I’ll be honest we didn’t fully appreciate why it mattered until we looked at the research. A Harvard field experiment by Leibbrandt & List found that when salaries were unclear, men applied more frequently and were far more likely to negotiate for higher pay, while women were more likely to accept the first offer. Simply stating the salary range significantly reduced that gap. It’s a perfect example of how a seemingly small detail can quietly reinforce inequality unless you look at it with a critical mindset.

We also changed how we assess incoming CVs. Instead of letting hiring managers see them first, we moved to CV screening carried out by a third party, based solely on the skills and experience we say we’re looking for. It strips out bias early in the process and makes everything cleaner and fairer. Add to that some structured interview scoring and a more mindful approach to shortlisting, and you end up with a system that works better for everyone.

The other thing we learned is that accessibility isn’t just physical. We’re a digital business with hybrid working as standard, but that only works if people have what they need at home. That’s why we support employees financially to set up proper home-working environments, including any reasonable adjustments. It’s a simple way of levelling the playing field, but again, it matters. We’re proud to be a Disability Confident Employer.

And finally, you can’t just focus on today’s applicants. We spend time talking in schools and colleges because the earlier people see themselves in this industry, the more likely they are to choose it. If we want a broader talent pool tomorrow, we have to help build it today.

The bigger point in all of this is that improving diversity isn’t about big dramatic gestures. It’s about being willing to question assumptions, use the data, and accept that even well-intentioned processes can produce skewed outcomes. Once you start looking at the detail, it becomes very clear: if you want different results, you have to design for them. I’m pleased to say the changes we made a couple of years ago have had a really good impact in achieving a more diverse and engaged workforce, and encourage anyone else to look at their processes through a critical eye and ask ‘what does the evidence actually tell us?’.