how to evaluate independent director candidates: test judgment rather than polish

Pilot the matrix on two anonymised historic profiles before launch. If every interviewer interprets financial expertise, independence of mind or scale differently, refine definitions and examples. The test can reveal criteria that duplicate each other or reward the current board’s career path without predicting the mandate. Calibration should not produce one ideal personality; it should ensure that different evidence is measured against a shared decision need. Retain the calibration note so later score changes can be explained rather than appearing designed for the preferred person.

The NRC should convert the role specification into a weighted evidence matrix before interviews. Separate essential committee or sector capability, future-board relevance, independence, integrity, capacity, motivation and learnable gaps. Define what strong evidence looks like for each. This prevents a charismatic conversation or promoter referral from changing criteria after the fact. It also allows different career paths to demonstrate the same underlying judgement without identical titles. The matrix should identify source and confidence for each score, distinguishing interview assertion, verified record, reference evidence and an assumption still awaiting diligence.

Use non-negotiables sparingly and explicitly. Failed statutory independence, material integrity concerns or impossible capacity can stop a candidacy; absence of prior listed-board service may be developable. Do not average a serious conflict against high sector expertise. Scorecards organise evidence but cannot remove judgement. The committee should record why a threshold exists and which missing facts require further diligence rather than translate unknowns into low scores automatically. Non-negotiables should be reviewed by counsel or the appropriate owner before interviews so a misunderstood legal criterion does not exclude the entire market incorrectly.

Assign interviewer roles. One person can explore committee depth, another board boundary and dissent, another motivation and time, with company secretarial support on eligibility. Common core questions make comparison fair, while follow-ups test the individual evidence. Avoid an unstructured panel where the most senior interviewer dominates and candidates are judged mainly on rapport with the chair. Independent interview notes completed before group discussion reduce hindsight editing after the chair announces a preferred interpretation of the candidate’s answer publicly to colleagues.

Ask for a situation with incomplete evidence, competing stakeholders and material downside. Explore what the candidate knew, which assumption mattered, what they asked, who owned action and what happened. Strong candidates acknowledge uncertainty, team contribution and a decision that did not work. Weak answers often claim sole success, avoid contrary evidence or describe operating execution without showing how board judgement would differ. Ask what evidence would have changed the candidate’s recommendation, revealing whether the answer is genuinely conditional or a story reconstructed around a successful outcome.

Add one company-relevant hypothetical without expecting free consulting. An audit candidate might examine a provision or revenue signal; an NRC candidate a succession conflict; a risk candidate a cyber or safety escalation. Assess questions before answers. The purpose is not to find someone who guesses management’s preferred solution, but to observe whether the person identifies authority, evidence, conflict, stakeholder and follow-up under time pressure. Different interviewers can score question quality, governance boundary and communication separately, producing richer evidence than one overall impression of confidence.

Section 149(6) and Regulation 16 for listed entities create legal criteria that require factual diligence. Independence of mind is behavioural: willingness to question a sponsor, change a view with evidence, disclose conflict and accept collective decisions. A person can pass the definition yet defer habitually to power. Interview references and decision cases should examine dissent without romanticising constant opposition. Constructive independence improves decisions rather than performing disagreement. References should include someone who observed the candidate under a powerful sponsor, because easy dissent in a low-stakes setting may not predict board independence.

Motivation reveals pressure points. Ask why this company, what the candidate expects to contribute, what would cause resignation and how fees fit personal economics. Prestige, networking or future consulting as primary motives can compromise boundaries. Candidates may reasonably value remuneration and learning; the committee should assess whether they could still challenge a promoter or lose reappointment. Avoid intrusive personal finance questions and focus on material dependence and disclosed relationships. If fees are economically significant, discuss role boundaries and renewal pressure rather than request intrusive details that are unnecessary to assess dependence risk.

Capacity should use actual calendars and crisis scenarios. Verify executive role, boards, committees, travel and results peaks under current directorship limits and employer permission. Ask how the candidate would handle two simultaneous urgent decisions. Attendance history is relevant but not conclusive. A candidate can appear at every meeting while underprepared, or miss one meeting for a legitimate conflict while contributing deeply across the year. Capacity evidence should include portal-reading and preparation blocks, not assume a free diary entry means the candidate can absorb technical papers.

Reference questions should follow the evidence matrix: preparation, financial or sector judgement, confidentiality, dissent, influence, conflicts and response to failure. Speak with people who observed different contexts, with candidate consent. A reference supplied by a close sponsor may still be useful if the relationship is known. Compare examples and investigate material inconsistency rather than count positive adjectives. Reference discrepancies should be put to the candidate with enough specificity for response while protecting lawful confidentiality and the referee’s legitimate privacy interests.

Background diligence should verify identity, chronology, qualifications, offices, litigation, regulatory matters, public conduct and relevant conflicts proportionately. Distinguish allegations, proceedings and findings and allow factual correction. Do not collect unrelated personal or family data. A discrepancy in title may be innocent or material depending on whether it inflated statutory authority. The NRC should receive supported conclusions and limitations, not raw internet results. Background reports should show source, date and limitation, preventing an old or mistaken online result from becoming an unchallengeable integrity label.

Before the NRC votes, ask each member to state the strongest evidence for an alternative and the greatest unresolved risk in the preferred candidate. This counters confirmation bias and shows whether advocacy is relationship-based. The chair can then summarise where evidence is clear, where judgement differs and which conditions or development will be recorded. A rigorous final discussion need not demean candidates; it protects them from a process where familiar support silently outweighs facts and gives the board a defensible reason for its recommendation.

The NRC should compare evidence while several viable candidates remain, identify development plans and state why the recommendation fits the future board. Relationships with committee members must be disclosed. A high score should not conceal a non-negotiable failure, and a narrow lower score should not override strong decision evidence if the difference reflects a learnable gap. Record the reasoning sufficiently for board and member materials. The recommendation can compare the preferred candidate with capability patterns rather than naming every unsuccessful person in broadly circulated board or member materials.

Communicate process status accurately and protect data. Do not imply appointment before approvals or keep people indefinitely without purpose. After appointment, test evaluation against the original evidence and revise criteria when predictions were wrong. This page is general candidate-evaluation guidance, not legal or background-check advice. Apply current company, listing, sector, employment, privacy and anti-discrimination rules, with the company retaining full responsibility for its decision. A post-appointment review should ask which interview evidence predicted behaviour and which criteria proved irrelevant, improving the next succession cycle.