From assumptions to evidence: what we learnt from testing what works in improving diversity data

There’s a phrase we use a lot at Youth Futures Foundation: if you’re not assessing, you’re guessing.

It gets to the heart of why evidence matters. Employers make decisions every day about how to recruit, retain and support young people, but sometimes approaches that feel intuitive or are widely regarded as good practice do not have the impact we expect.

Diversity data is key to this. Good workforce data helps employers understand who is in their organisation, where inequalities exist and where action is needed. But its value depends on employees choosing to share their information. So how employers encourage disclosure matters, and it’s an area where we need evidence about what actually works.

That’s why I’m really excited that Youth Futures has completed this Randomised Controlled Trial (RCT), working with Kier Group and the Behavioural Insights Team to test whether changing the way young employees are asked to disclose their diversity data can make a difference. It is the first in a wider programme of trials through which we are partnering with employers to test practical approaches to recruiting, retaining and supporting young people.

Delivering rigorous RCTs, or ‘trials’, in real-world employment settings is challenging, and we’re immensely grateful to Kier for being such a pioneering partner: willing to put its own practice to the test, learn from the results, and share those lessons with others. It is this kind of leadership and willingness to innovate that will help us understand what really makes a difference for young people and give other employers the confidence to act.

More than 9,000 Kier employees were randomly allocated to receive one of two email campaigns asking them to share information across nine diversity characteristics, including disability, neurodiversity, ethnicity and social mobility.

One group received a ‘standard’ communication with detailed information about why Kier was collecting the data and how it would be used. The other received a ‘behaviourally informed’ communication. It was shorter, more visual and personalised, came from a named senior leader and included a prominent call to action with a time estimate for completing the request.

What did we find?

The result was not what we expected.

Among young employees, the standard communication was more effective than the behaviourally informed version at encouraging them to disclose all nine diversity characteristics: more than one in four receiving the standard communication disclosed all nine compared with around one in six receiving the behaviourally informed version.

This matters because our assumption, based on lower quality existing studies, had been that the behaviourally informed approach would be more effective. And that, for me, is exactly why trials like this are so important. We tested an assumption and the evidence told us something different.

The qualitative research alongside the trial helped us understand why:

• Purpose and trust emerged as important factors. Employees wanted to understand why Kier was asking for their diversity information and how it would be used. Those receiving the behaviourally informed communication, which contained less detail, wanted more explanation of its purpose beyond the individual benefits to them.
• Employees responded positively to the brevity and personalisation of the behaviourally informed communication, but these preferences didn’t translate into greater disclosure.
• The messenger and method also mattered: a named senior leader may not be equally familiar to everyone, while a prominent call-to-action button may also have raised cybersecurity concerns.
• Ease of the process was important too. Employees in both groups were given a direct link to a simple form, rather than having to navigate the usual HR system.

Taken together, the findings show that diversity disclosure is an organisational challenge, not simply a communications issue. How you ask matters, but so does whether employees understand the purpose, trust how their information will be used, and find the process straightforward.

Putting evidence into action

There is no single winning formula here – what works in one workplace cannot be assumed to work everywhere. The bigger lesson is that employers need to be willing to test, learn and adapt. We have put together an employer summary guide which pulls out practical lessons from the trial for employers. Early next year, we will publish further practical support to help employers apply and adapt the learning in their own workplaces.

And this trial is only the start. Over the next two years, we’ll share findings from more trials testing everything from work experience and line management of apprentices to mental health support at work. Together, they will help us build a much clearer picture of what employers can do to help young people get into work, stay in work, and thrive.

Tackling the youth employment challenge requires more than good intentions. It requires us to be curious about what works, honest when the evidence challenges our assumptions, and willing to keep testing and learning together.

If we’re not assessing, we’re guessing. And for the young people who will make up the workforce that every sector depends on in the years ahead, we need to do better than guess.