Diversity Hiring at Leadership Level: How Slates Are Built
Diversity hiring at leadership level is a sourcing discipline, not a quota. How diverse shortlists are actually built, and what the evidence says works.
Diversity hiring at leadership level means widening the population you research before anyone is approached, then assessing every candidate against the same evidence standard. It is not a quota applied to a shortlist. The decisive research finding is blunt: with only one woman among the finalists, the statistical odds of her being appointed are effectively zero.
What is diversity hiring at leadership level?
At CXO and board level, diversity hiring is a sourcing and assessment discipline rather than a hiring preference. The work happens at mapping stage. You build a deliberately wider research universe, adjacent sectors, people one level below the title who run equivalent scope, returners whose service was interrupted, and then you assess everyone against the same structured brief. That is what a diversity search mandate is actually buying.
What it is not: adding a name late to balance a list. Candidates notice, panels notice, and the person added late is rarely in a position to win. That is worse than doing nothing, because it confirms to the exact population you are trying to reach that the process was decorative.
Why does one woman on a shortlist change nothing?
Because a lone candidate reads as a category rather than a contender. Stefanie Johnson, David Hekman and Elsa Chan examined 598 job finalists and published the result in Harvard Business Review in 2016. When at least two women were in the finalist pool, the odds of a woman being hired were 79.14 times greater than when there was only one. For minority candidates, the odds were 193.72 times greater with two in the pool.
The mechanism is status quo bias. A single different candidate is measured against the type. Two make difference itself unremarkable, and the panel goes back to comparing people on evidence. The implication for a brief is precise: a slate carrying one woman among five is, statistically, a slate carrying none.
How is a diverse slate actually built?
Four things, in this order.
- Test the brief against the market first. Most briefs carry three or four requirements nobody has checked. A named competitor set, a particular institution, unbroken service, each reads as a quality standard and each removes populations for reasons unrelated to capability.
- Map the full population rather than the network. Referral sourcing is homophilous by construction: a search run through the hiring executive's contacts returns people like the hiring executive, reliably, without anyone intending it. Proper market mapping is the only stage at which breadth can still be created.
- Search one level down and one sector across. Where representation at the title is thin, insisting on someone who already holds that title guarantees the shortlist reproduces the existing distribution, you are recruiting from precisely the layer that has the problem.
- Assess everyone identically. Same structured questions, same evidence standard, same written record. Unstructured interviewing is a weak predictor of senior performance for anybody, and the noise it introduces does not fall evenly across candidates.
Three of those four happen before a single candidate is approached. If the researched population was narrow, no amount of care at shortlist stage recovers it, which is why diversity is decided in weeks two and three of a standard eight-stage search, not at the point the list is presented.
What does the India data actually show?
Women hold 19 per cent of C-suite roles in India against a global average of 30 per cent, according to an Avtar Group study reported in January 2025. In the same study, 44 per cent of respondents identified bias in hiring and promotion as a barrier, and 41 per cent cited a shortage of qualified female candidates for leadership roles.
Those last two findings sit awkwardly together, and that tension is the whole subject. A shortage at the exact title is usually real. A shortage of capability at equivalent scope usually is not. The distance between the two is what a wider map closes, and it is the only part of the problem a search firm can do anything about.
Attrition compounds it. The same study put senior-leadership attrition for women at 8 per cent in 2024, improved from 10 per cent in 2020, but still double the 4 per cent recorded in 2019. Hiring into a leaking pipeline is expensive and the arithmetic does not improve with repetition.
The Women Leadership Survey 2026, published in February 2026 by AIMA and KPMG in India, locates the pressure point at mid-career, where attrition risk is highest and leadership momentum most often stalls. Its finding on cause is the one worth carrying into a hiring plan: exits are rarely driven by declining ambition. They reflect accumulated organisational friction, role design and uneven support through transitions.
Is there a commercial case, or only a compliance one?
Both, and they are separate arguments that should not be run together. The compliance floor in India is statutory: the Companies Act and SEBI's listing regulations set minimum board composition, which we set out in detail in our analysis of women on boards in India. A floor is not a strategy. Compliance produces exactly the appointments compliance requires, and nothing beyond them.
The commercial case stands independently. McKinsey's Diversity Matters Even More, published in December 2023, found companies in the top quartile for gender diversity on executive teams were 39 per cent more likely to outperform financially, up from 15 per cent in 2015. Companies in the bottom quartile on both gender and ethnic diversity were 66 per cent less likely to outperform.
Treat correlation as correlation. The honest reading is not that appointing a woman raises margin. It is that organisations able to recruit from the entire market, rather than a quarter of it, tend to be organisations that are well run in other respects too. That is still a reason to widen the map, but it is a different reason from the one usually offered.
Does a wider slate mean a lower bar?
It means a larger denominator. A shortlist of five drawn from a researched population of four hundred is a different object from a shortlist of five drawn from whoever replied, and only the first survives interrogation by a board. That is the same argument that makes a search worth running on retainer rather than on contingency.
The quality risk in fact runs the other way. Title matching and referral sourcing feel rigorous and are not: they narrow the field on proxies rather than on evidence. Structured assessment applied to a wider population raises the standard for every candidate, not only the under-represented ones.
What should not happen is a fixed proportion imposed at shortlist stage. Presenting a candidate who cannot win the role wastes that person's time, damages the client's standing with the population it is trying to reach, and is always discovered. Every name on a list should be there because they could be appointed.
What should a board ask its search firm?
- How large was the researched population, and who was rejected and why? A shortlist without the names behind it cannot be interrogated.
- Which requirements in our brief did you test against the market, and which narrowed the population most?
- How many candidates from under-represented groups reached final stage, rather than first conversation? Pipeline statistics flatter; finalist statistics do not.
- Is the assessment structured and written down, or a series of conversations followed by a view?
- What is your position if we ask you to guarantee a proportion on the list?
A firm that answers the last question by simply agreeing has told you what its shortlists are worth. The answer you want is that breadth is guaranteed at mapping stage and outcomes are not guaranteed at all, coupled, where the pipeline is genuinely thin, with assessment and succession work that builds the bench rather than repeatedly searching an empty one.
What a search cannot fix
Recruitment does not solve retention. If appointments join and leave inside two years, the search was not the failure. The question is who gets sponsored, heard and promoted once inside, and it is considerably cheaper to answer that before commissioning another search than after.
Nor can a search compress a market. Where four people in the country have run the function at that scale and two are unmovable, the answer is a different brief, usually one level down or one sector across, not a longer list. Our analysis of why executive hires fail reaches the same conclusion from the opposite direction, and it applies to every senior mandate we run, diverse or otherwise.