Evidence-based HR starts with a better question

“Onboarding takes too long.” It sounds like a clear problem. But ask a few more questions and you may hear three different accounts: equipment arrives late, managers have not agreed on priorities, or new employees need more practice than the schedule allows.

Each calls for different work. Buying a new system before understanding the delay could leave the team with the same problem and another tool to maintain.

Evidence-based HR helps you make a more informed choice. CIPD’s evidence-based HR guide brings together research, organizational data, professional expertise, and stakeholder perspectives. It describes a process of asking, acquiring, appraising, combining, applying, and assessing evidence. The aim is to strengthen judgment by considering several sources and their limitations.

Make the decision clear

Before asking for a proposal or commissioning a project, write a short brief. Keep it focused on what you need to decide.

• Describe what you see. “Access is regularly missing on a new hire’s first day” gives you something to investigate. “Our onboarding is broken” leaves too much open.
• Name the choice. Are you deciding who owns a handoff, whether to add support, or whether to investigate further?
• Identify the people involved. Include those doing the work and those receiving it. Their accounts may differ for good reasons.
• Define useful progress. Choose an outcome that matters, then consider whether measuring it could encourage shortcuts or move work onto another team.

Check the explanation before choosing the fix

For the onboarding example, follow a few recent hires through the step causing concern. Compare planned and actual dates. Ask employees where they waited, managers what they expected, and coordinators what they needed to move the work forward. Check whether the experience differs by role or work arrangement.

Use that understanding to guide your research. Look beyond a headline or an impressive statistic. Who participated? What was compared? What was measured? How closely does that setting resemble your workplace? A study showing an association can point to a useful question without establishing what caused the outcome.

Evidence should help you examine your preferred explanation, including the possibility that it is incomplete.

Make a recommendation you can review

A recommendation might read: “The records and interviews suggest that unclear ownership contributes to access delays. We propose testing a named owner and an earlier readiness check.” This is an example of how to frame a proposal, rather than a client case.

Add what remains uncertain, who will make the change, and when you will review it. Check whether the new step was used, whether the delay changed, and whether another team absorbed extra work. Collect only the information the review needs and handle personal information appropriately.

Report what happened with care

A before-and-after comparison can help you learn, but workload, different roles, seasonality, or other changes may also explain the result. Keep those limits in the account.

Your next HR decision can begin with a single page: the question, the evidence, the uncertainty, and the review plan. That gives everyone something concrete to challenge, act on, and learn from.