What Is Evidence-Based Hiring?
The phrase gets used often enough that it's worth actually defining, rather than assuming everyone means the same thing by it.
The simple version
Evidence-based hiring means making a hiring decision based on what a candidate has actually demonstrated, not what they've claimed on a CV, how they performed in a single unstructured interview, or how closely they happen to resemble the person interviewing them. It's the deliberate application of real, verifiable signal in place of assumption.
What it isn't
It isn't a synonym for "using AI." AI can be applied to evidence-based methods, and it can equally be applied to methods that aren't evidence-based at all, an AI tool that just re-ranks CVs by keyword match hasn't introduced any more evidence than a human doing the same thing manually. The label describes the kind of signal being used, not the technology doing the processing.
It also isn't the same as "data-driven." A hiring process can generate plenty of data, interview scores, personality test results, without any of that data being genuinely predictive of job performance. Evidence-based hiring specifically means using signal that's been shown, through real research, to actually predict how someone will do the job.
Where the term actually comes from
The underlying idea has a long research history in industrial psychology, decades of study into which selection methods genuinely predict job performance, and which ones feel rigorous but don't. That research consistently ranks methods by their validity, essentially, how strongly a method's results correlate with actual on-the-job performance. Traditional unstructured interviews and years of experience alone rank surprisingly low. Structured, realistic work samples rank near the top.
Why it matters now more than ever
CVs have become easier than ever to polish, and increasingly easy to generate convincingly with AI assistance, which means the CV was already a weak signal and is becoming weaker still. At the same time, unconscious bias in traditional interviews is well documented: candidates who happen to resemble the interviewer are consistently rated more favourably, regardless of actual capability. Evidence-based hiring exists specifically to correct for both problems at once, by grounding a decision in something a candidate actually did, observed directly, rather than something they said about themselves or an impression formed in forty-five minutes.
What it looks like in practice
Concretely, evidence-based hiring usually means some form of realistic, structured assessment, a work sample, a simulation, a scenario built around what the role actually demands, evaluated consistently against the same criteria for every candidate. The result is a hiring decision grounded in observed behaviour, not inference.