Insights · AI & Recruitment

Reference checks: getting useful rather than formal feedback

Reference checks add the context of real collaboration to interview evidence. They should not be a search for damaging information or a formal date-confirmation exercise.

Choosing the right respondents

The most useful feedback comes from people who directly observed the candidate’s work: a manager, colleague or internal client. Proximity to the work matters more than seniority. For leadership roles, combine the view of a manager with someone who worked alongside or reported to the candidate.

Questions about behaviour and outcomes

Instead of asking whether someone is “good”, explore role context, strongest contribution, difficult situations, decision-making and response to feedback. Ask which environment supports the candidate’s best work and what support helps during onboarding.

Handling differences

A reference is one source, not a final verdict. Different people observe different parts of the work. Separate facts from interpretation, compare them with interview evidence and return to the candidate for a respectful clarification when needed.

Ethics and confidentiality

Never contact a current employer without explicit permission. Candidates should know the format and respondents. Reports should contain relevant professional information without unnecessary private detail or unsupported judgement.

Practical checklist

  • Candidate consent
  • Two relevant respondents
  • Consistent question structure
  • Facts separated from opinions
  • Comparison with role criteria

Frequently asked questions

Why are two references useful?

Two perspectives reduce dependence on one subjective experience and provide broader working context.

When should checks happen?

Usually at the final stage when there is mutual interest and the candidate has approved the contacts.

Related insights

Continue reading

TalentOne

Need to hire for a new way of working?

TalentOne can clarify the profile, assess the market and find a candidate in the context of AI transformation.

Submit a request