African Women CEOs SummitAn Ishmael Yamson & Associates initiative

The Top 100 Women CEOs in Africa

Selection
methodology.

A clear standard for executive responsibility. Evidence-led assessment. Independent decisions. Recognition that cannot be bought.

Adopted policy · Version 1.0

Adoption confirmed 18 November 2026

Unranked recognition · No nomination fee

An Ishmael Yamson & Associates initiative. With research support from Reset Global People.

01 / Selection policy

What inclusion means

The Top 100 recognises women whose chief executive authority and demonstrated leadership can be substantiated through a fair, independent process. Inclusion means a candidate meets the published eligibility and evidence standards and is selected through this assessment policy.

The List is unranked. It is not a claim that these are objectively Africa’s 100 best executives, a census of women CEOs, an investment recommendation or certification of an employer.

  • Nomination, consideration and official recognition require no payment. Sponsorship, tickets, Summit attendance and alumni participation have no influence on selection.
  • Actual executive responsibility matters more than a job title, celebrity or organisation size.
  • All candidates face the same published requirements, regardless of how they enter the process.
  • Historical honorees retain their original distinction. This policy applies prospectively and does not retrospectively reassess earlier editions.

The policy sets the standard for future selection cycles. An edition’s cut-off dates, supported languages, committee appointments and publication calendar are published before nominations open. Policy adoption is not a claim that pilot testing, appointments or a new assessment cycle have already been completed.

02 / Selection policy

Who qualifies for consideration

A candidate must be a woman serving as CEO, Managing Director, Group CEO or an equivalent chief executive of an identifiable qualifying enterprise. All four responsibility tests must be met together.

The four executive responsibility tests
TestRequired findingTypical evidence
AuthorityPrimary strategic and operational executive authorityBoard-approved mandate or governance documents
AccountabilityResponsibility for overall enterprise performanceReporting relationships and performance mandate
OrganisationLeadership of an identifiable enterprise, not only a functionLegal entity or documented operating business
VerificationA role and scope that can be credibly substantiatedCurrent official record plus corroboration

Organisational scope

Legally established businesses, regulated financial institutions, qualifying public enterprises, nonprofit organisations and operating institutions are eligible where an identifiable executive mandate and substantive operations exist. Elected political office alone, personal billing or a nominal company title does not qualify. There is no blanket revenue or employee minimum across sectors.

Titles, founders and shared authority

Founders qualify while exercising the required executive authority. Non-executive owners, chairs, board members and functional executives do not qualify solely by status. Executive chairs, presidents, partners and executive directors require a recorded finding against all four tests.

Co-chief executives may qualify individually where their shared whole-enterprise mandate is documented; each contribution is assessed separately. A subsidiary CEO can qualify at subsidiary scope even when reporting to a group CEO. Parent-group results cannot be credited without a documented connection. Distinct enterprises within one group may produce eligible candidates, with duplicated claims checked rather than an undisclosed company quota.

Geographic boundary

The adopted boundary is african citizenship and substantive african enterprise operations. Dual citizenship qualifies. Diaspora residence does not disqualify a woman leading the relevant African enterprise. Substantive operations mean verifiable activity and organisational responsibility in one or more African countries, not citizenship alone, a remote advisory role, a sales contact or an African parent brand. Operating locations and nationality are recorded separately.

Tenure and assessment period

At least 12 continuous months in the qualifying chief executive role are required by the edition cut-off, whether permanent or acting. Assessment covers the preceding 36 months, restricted to time in the relevant role, using the latest 2 available annual reporting periods where possible. Earlier career experience provides context, not credit for results achieved before assuming authority.

Role changes

Roles are rechecked before publication. Departure before publication normally results in deferral. A move to another CEO role requires documented review; unrelated appointments cannot be combined to satisfy tenure. After publication, a new role or retirement updates the current-role field without erasing edition recognition. There are no private founder exemptions.

03 / Selection policy

One standard, three entry routes

Self-nominations, third-party nominations and research-led identification enter the same register. Multiple nominations, public voting, social-media engagement and lobbying do not improve a score. Research can identify an eligible executive without a public nomination.

Continental research coverage

Research draws on official company records, exchange disclosures, regulator directories, credible sector bodies and checked nominations. Coverage spans Northern, Western, Central, Eastern and Southern Africa, language communities and sectors rather than only English-language publicity. Coverage targets guide research effort; they do not guarantee selection places. There are no country, sector or language quotas.

The process tracks gaps at discovery, contact, eligibility, evidence completion and selection. Translation and reasonable assistance address avoidable barriers without changing the evaluation standard.

Consent and candidate assistance

Intake records the nominator’s relationship, enterprise, role dates, operating locations and supporting sources. Candidate consent is required before requesting non-public evidence or publishing a profile. Third-party nomination does not authorise portrait use or access to confidential information.

Clarification support carries no charge. Equivalent documentary formats are accepted where they substantiate the same facts. Candidates receive 10 working days for material clarification, with one reasoned extension of up to 5 working days for access or translation needs. Assistance is logged consistently, and supported languages are stated in the edition calendar.

Previous honorees can enter a new cycle without automatic advantage. Current eligibility and achievements are reassessed. Records are deduplicated, and a person receives at most one recognition per edition. Candidate information is never used to sell tickets, services or recognition products.

04 / Selection policy

Evidence before assessment

Every eligibility finding and material claim supporting a criterion must be substantiated. Preferred sources are official filings, audited or professionally prepared reports, regulator records, governance documents and direct institutional confirmation. Independent reporting corroborates events and context; candidate narratives remain claims until supported.

Role and enterprise scope require an authoritative institutional source plus a distinct corroborating record or direct research verification. Each decisive performance claim records its basis and corroborating source or verification procedure. Two articles repeating one press release count as one underlying source.

Evidence status is separate from score
StatusMeaningAssessment use
VerifiedSupported and checked material factMay support assessment
QualifiedSupported with a recorded scope or comparability limitUsable only within that stated limit, with reasons recorded
UnverifiedInsufficiently substantiated assertionCannot justify a positive score
ContradictedReliable records materially disagreeResolve the conflict or defer the decision

The evidence gate passes only when all eligibility findings and all five criteria are assessable. Missing core evidence is “not assessable,” not zero; weights are never redistributed to hide a gap.

Private evidence and precise claims

Private enterprises may provide redacted accounts, confidential accountant confirmation or documented source inspection with permission. Limited public financial data is not itself poor leadership, but confidentiality does not remove substantiation requirements.

Quantitative claims record periods, units, currencies, source dates and enterprise scope. Revenue is distinguished from profit, assets from revenue, forecasts from achieved results and beneficiaries from unique beneficiaries. Currency comparisons require explicit conversion dates and sources. Conflicting figures are reconciled or excluded, and estimates remain estimates. AI summaries, search snippets and biographies can locate sources but cannot provide final material verification.

05 / Selection policy

Five dimensions of executive leadership

Executive contribution and performance together account for 55% of assessment. Organisational significance contributes 10%; responsible leadership is both a scored dimension and a minimum safeguard. These weights express policy priorities, not externally validated constants.

A

Executive contribution

25%

Consequential decisions, strategic execution, organisational change and team leadership during the assessment period. Evidence must connect the candidate’s actions and responsibilities to outcomes; testimonials alone do not establish causation.

Criterion-specific scoring anchors
  1. Competent documented decisions, with limited evidence of distinctive contribution.
  2. A clear strategic decision or execution programme with a substantiated contribution to positive results.
  3. Sustained delivery across consequential decisions, corroborated personal responsibility and strong team execution.
  4. Unusually consequential leadership in demanding conditions, multiple attributable outcomes and a credible comparison with the starting position.

Scores 0 and 1 follow the common scale. NA means insufficient evidence, not poor performance.

B

Enterprise performance and resilience

30%

Sustainable growth, financial health, service delivery, turnaround progress and risk management are assessed against the enterprise’s starting position, credible targets and market conditions. Growth accompanied by deteriorating solvency or unresolved operational harm is not automatically strong performance.

Criterion-specific scoring anchors
  1. Broadly adequate mandate delivery, with material operational or financial limitations.
  2. Credible improvement or sustained strong performance against relevant baseline and risk conditions.
  3. Sustained high-quality results across several relevant indicators, with external factors explained.
  4. Exceptional evidenced improvement, resilience or mandate achievement in demanding context, without concealing adverse indicators.

Scores 0 and 1 follow the common scale. NA means insufficient evidence, not poor performance.

C

Responsible leadership and governance

20%

Effective controls, fair workforce practices and the response to material risks or failures. Basic compliance is not proof of exceptional leadership, and serious substantiated misconduct cannot be offset by commercial success. Absence of reported scandal is not evidence of strong governance.

Criterion-specific scoring anchors
  1. Basic governance and accountability exist, but important weaknesses remain.
  2. Effective controls, documented accountability and a credible response to material risks.
  3. Sustained governance quality, fair workforce practices and improved risk stewardship.
  4. Unusually strong accountable leadership, demonstrated prevention or remediation and institution-wide adoption of effective practices.

Scores 0 and 1 follow the common scale. NA means insufficient evidence, not poor performance.

D

Wider economic or societal contribution

15%

Verified employment quality, productive investment, essential-service access, innovation, supplier development and other relevant outcomes. Demonstrated results matter more than SDG labels, charitable publicity or announced commitments. Outcomes already credited elsewhere are not counted again.

Criterion-specific scoring anchors
  1. Verified useful contribution, with limited breadth, depth or measurement.
  2. Clear stakeholder outcomes with a credible link to enterprise activity.
  3. Sustained meaningful outcomes supported by consistent measures and corroboration.
  4. Unusually substantial outcomes in context, with strong evidence of reach or depth and the enterprise’s contribution.

Scores 0 and 1 follow the common scale. NA means insufficient evidence, not poor performance.

E

Organisational responsibility and significance

10%

The candidate’s actual accountable enterprise, its complexity and sector-specific significance. Revenue and headcount provide context, not automatic points. Wealth, fame, career prestige and the scale of an unrelated parent group do not determine the assessment.

Criterion-specific scoring anchors
  1. Substantive enterprise responsibility of relatively limited complexity or contextual significance.
  2. Meaningful strategic, operational and stakeholder complexity.
  3. Substantial accountable scope, sector importance or essential-service responsibility.
  4. Unusually consequential organisational responsibility, supported by actual enterprise scope and operating context.

Scores 0 and 1 follow the common scale. NA means insufficient evidence, not poor performance.

06 / Selection policy

Common criteria. Relevant context.

Each file uses the same five criteria, scale and weights, with indicators appropriate to its operating model. A context note records baseline, sector conditions, comparator basis, evidence availability and external factors. This is structured expert judgment, not precise statistical equivalence.

Illustrative sector indicators
OrganisationRelevant indicatorsImportant limitation
Financial institutionAsset quality, capital resilience, access and returnsSize and profit omit risk and regulation
Industry or infrastructureReliability, productivity, safety, investment and financial healthCommodity prices can dominate results
Technology or growth enterpriseRetention, unit economics, quality and executionFunding raised is not revenue or sustainable performance
Public or nonprofit institutionMandate outcomes, service quality, stewardship and controlsProfit is not a universal measure of success
Professional or service enterpriseClient outcomes, workforce development, quality and sustainabilityA partner title alone does not establish CEO authority

Sector guidance is developed before live assessment. Reliable regulator, industry or official comparisons are used only when definitions and periods align. Reviewers cannot invent peer percentiles, convert incomparable outcomes into a league table or make population-wide claims from small, biased datasets.

Each outcome has a primary scoring purpose. Revenue growth can support performance while evidence of the executive’s decisions supports contribution, but the same raw figure is not fully credited across multiple criteria. Coverage gaps are addressed through research and access, not hidden bonuses or late geographic preferences.

07 / Selection policy

Scoring and minimum safeguards

The common scoring scale
ScoreAnchorInterpretation
0Demonstrated failureEvidence supports material failure for the criterion.
1WeakSubstantial weaknesses with limited positive delivery.
2AdequateCompetent delivery with material limitations.
3StrongClear, substantiated positive delivery in context.
4DistinguishedSustained high-quality delivery with corroborated contribution.
5ExceptionalUnusually strong evidenced achievement against credible context.
NANot assessableInsufficient evidence: hold the file rather than treat missing evidence as poor performance.

Reviewers assign whole-number initial scores. A 5 requires a specific explanation of why the evidence exceeds the 4 anchor. Large-company status alone never justifies a high score.

How the total is calculated

Each criterion contributes its percentage weight multiplied by its score, divided by 5. The weights total 100, so the maximum total is 100. Final criterion values are the mean of two valid ratings, or the median of three valid ratings after a third review. The final total is recalculated from those values; moderated and unmoderated totals are not mixed.

  • A final total of at least 70 out of 100.
  • Every criterion value must be at least 2.
  • Responsible leadership and governance (C) must be at least 3.
  • All eligibility and evidence gates must pass, with no unresolved material integrity issue.

These gates qualify a candidate for consideration; they do not create an entitlement to one of the 100 places. Exact arithmetic is retained for comparison. Internal totals are displayed to one decimal, but rounding never creates a pass or resolves a tie. Individual scores and ordinal positions are not published.

An illustrative calculation

A hypothetical candidate with final values A 4, B 4, C 4, D 3 and E 5 receives 20 + 24 + 16 + 9 + 10 = 79 points. She meets the numeric gates, but still needs integrity clearance and final selection. A governance score of 2 fails the governance floor despite a competitive total. If B is NA, no valid total exists.

08 / Selection policy

Independent review and final selection

Each assessable file goes to 2 independent, non-conflicted reviewers, including at least one with relevant sector competence. Balanced workloads, overlapping assignments and calibration help detect systematic leniency or severity. Before live assessment, 10–15 practice dossiers test sectors, regions, evidence conditions and borderline roles.

When a third review is required

  • Any criterion ratings differ by 2 or more points.
  • Reviewer totals differ by more than 10 points.
  • A reviewer gives NA, or reviewers disagree on a material evidence or integrity finding.

NA and material fact conflicts must be resolved before aggregation. The median is used only after all three criterion ratings are valid. Original scores remain on record; revisions are attributed to the reviewer with dates and reasons. The committee cannot add undocumented points or substitute a preferred total.

The selection boundary and ties

Qualified candidates are ordered internally by final score. Boundary ties are resolved by higher B, then A, then C. If still tied, every tied file must have three independent valid reviews; files already reviewed three times receive no fourth reviewer. Criterion medians, full ordering and tie rules are recomputed.

If a tie remains, the non-conflicted committee records a comparative judgment using only published criteria and existing evidence. Each final inclusion requires two-thirds support among voting members. Written reasons explain the distinction. Sponsor preferences, celebrity and geographic preference are prohibited tie-breakers.

The committee ratifies the slate and evidence gates. Departures from score ordering require a documented factual correction, integrity outcome or adopted tie procedure, with affected comparisons rerun.

Shortfalls and reserves

If fewer than 100 candidates qualify, research or verification is extended under a publicly revised calendar without lowering standards. A persistent shortfall defers the full Top 100 edition; a smaller, differently named recognition requires a separate institutional decision.

An ordered, cleared reserve list follows the same rules. Before publication, a withdrawal or ineligible selection is replaced by the next cleared reserve with reasons recorded. After publication, names are never quietly substituted: material corrections and replacements require public explanation.

09 / Selection policy

Independence, conflicts and integrity

The initiative is led by Ishmael Yamson & Associates, with research support from Reset Global People. Prospective eligibility and inclusion are decided by the independent selection committee. Founder and commercial staff cannot add candidates, suppress concerns or require recognition of a sponsor’s executive.

The committee has 5–7 members with sector expertise and continental coverage. Final deliberations require at least 5 non-conflicted members; suitable alternates replace recusals before deliberation. Research and administrative staff advise without voting unless appointed to a permitted independent role.

Disclosure and recusal

Members disclose employment, advisory, financial, close personal and family relationships, as well as relevant rivalries. Professional relationships have a 3-year lookback; older or indirect relationships are disclosed when material. Conflicted members do not review, discuss, vote or hear an appeal for that candidate, and receive no unnecessary private material.

Committee members, the founder, direct programme staff and paid selection contractors are ineligible in the cycle they serve. Historical recognition is unaffected. Sponsor affiliation alone does not disqualify a candidate, but confers no advantage.

Fair handling of integrity concerns

Credible regulator findings, court records, audited disclosures and substantiated reporting are considered for relevance, reliability, timing, individual responsibility and remediation. Allegations are not proven facts. Political criticism, commercial rivalry and anonymous accusations alone do not establish misconduct.

A non-conflicted integrity panel considers credible material concerns. The candidate receives the substance of the concern, within lawful confidentiality limits, and 10 working days to respond. An unresolved issue holds candidacy without publicly implying guilt. Adverse eligibility findings have recorded reasons and an appeal route.

No commercial influence

There are no nomination processing fees, bought awards, paid score upgrades or preferential official packages. Tickets, optional experiences and sponsorship negotiations are separate. Attempts to exert commercial influence are logged and escalated. Non-attendance is never treated as withdrawal from recognition.

10 / Selection policy

Publication, recognition and continuity

Provisional selections receive a factual check of name, role, enterprise, edition eligibility and profile facts, with a 10-working-day response period. Inclusion remains subject to final clearance. Non-response cannot create consent or justify invented details. Profiles use a consistent structure and cannot be lengthened through payment or have editorial reasons controlled by candidates.

What the public record contains

  • Methodology version, edition calendar, eligibility boundary, assessment period, criteria and weights.
  • Committee composition where appropriate, conflict safeguards, no-payment policy and correction route.
  • Deduplicated process counts, with nominations and research-led identification reported separately and broad categories protecting confidential candidates.
  • Material research or evidence limitations, without implying exhaustive continental coverage.

The List is published alphabetically with clear edition years, not ranked first to hundredth. Each honoree receives the same official distinction, digital certificate, verified editorial profile and eligibility for the core Alumni Network without purchasing membership. Alumni participation requires separate consent. Physical recognition logistics do not create paid superior honoree status.

The Annual Summit is separate

The Annual Summit is confirmed for March 2027 · Accra, Ghana. Invitations, tickets, travel and hospitality follow a separate process and do not affect selection or recognition. An honoree is not described as a confirmed delegate without that commitment. Summit timing does not override the safeguards or research timetable of a List edition.

The archive preserves original edition decisions and material amendments. Current job changes are recorded separately. Historical recognition remains under its original framework, and repeat honorees are deduplicated in totals of individual leaders.

11 / Selection policy

Corrections, appeals and complaints

Three distinct review routes
RouteWhat it addresses
Factual correctionNames, roles, enterprise scope and documented profile information
Eligibility or procedural appealMisapplied rules, timely evidence overlooked, conflicts or material process errors
Misconduct complaintIntegrity concerns or interference with selection

Assessed candidates receive their disposition and concise reasons consistent with confidentiality. Appeals are free and must be lodged within 10 working days of notification. Candidates receive their own eligibility findings and sufficient criterion feedback, without other candidates’ records or unnecessary disclosure of confidential reviewer identities.

A panel of 3 non-conflicted reviewers who did not assess or decide the case targets acknowledgment within 3 working days and a reasoned determination within 20 working days. Extensions for additional evidence or translation are communicated. Notification and appeals precede publication sufficiently to allow a meaningful remedy.

An appeal can uphold, correct, remit for reassessment or overturn an eligibility or procedural finding. It never guarantees inclusion. Changed scores or eligibility require affected boundary comparisons to be rerun and relevant candidates notified.

  • Disagreement with good-faith comparative judgment alone is not an appeal ground.
  • New achievements after the cut-off belong to the next cycle. Fresh evidence may substantiate a timely submitted fact or resolve a material error, but is not a private late application route.
  • Minor factual errors are corrected promptly. Material post-publication errors receive independent review and an opportunity for the honoree to respond; withdrawal or replacement requires a proportionate public explanation.
  • Later misconduct does not automatically rewrite an earlier assessment. Retrospective withdrawal requires a separately adopted fair policy; current network conduct is a separate governance matter.
  • Death, incapacity, restructuring or disputed authority is handled using published dates and evidence rules. Material policy gaps pause the affected decision for prospective clarification, not an invented honorary place.

To request a correction, raise a procedural concern or submit an appeal, email partners@africawomenceosummit.com with the edition, person or record concerned, the review route and supporting evidence. Do not send sensitive identity documents in an initial enquiry.

12 / Selection policy

Confidentiality and process assurance

Only information necessary for eligibility, assessment, communication and verification is collected. Public profile permission, private evidence access and alumni consent remain separate. Unnecessary identity documents are not required; citizenship records are not published. Necessary verification uses the least intrusive reliable method.

Candidate files are separate from sponsor and sales systems. Reviewers access assigned material only, under confidentiality obligations, with access recorded where feasible. Translators and specialist verifiers receive equivalent safeguards. Private files must not remain in personal accounts after assignments end.

Retention and lawful holds

Final decisions, public edition records and corrections form the institutional archive. The private-evidence default for non-selected candidates is deletion or irreversible de-identification 24 months after cycle closure. Selected-candidate private evidence is retained for 5 years, then reviewed for continuing need. Applicable duties and processing arrangements must be confirmed before collection.

Complaints, appeals or lawful preservation duties may require a controlled hold with a recorded basis and review date. These defaults are not a declaration that a compliant processing arrangement already exists in every jurisdiction.

Independent process assurance

Before publication, an independent process reviewer examines all 100 selected eligibility and evidence-gate records, every conflict or integrity exception and a risk-based scoring sample. Boundary cases, material moderation and a sample of non-selected files are checked for consistent application. Unresolved material defects stop publication.

The scope of review is stated publicly. A procedural review is not certification of every enterprise result; an internal review is never described as an external audit. Aggregate findings track evidence completeness, reviewer disagreement, clarification time, recusals, appeals, upheld errors and coverage by region, language and sector.

13 / Selection policy

Edition timing and readiness gates

The sequence below is relative to a List edition’s publication date, not an announced nomination timetable or a confirmed 2027 List publication date. Each edition calendar sets cut-offs, working-day conventions, holidays and time zones.

  1. 9–6 months before publication

    1. Readiness and calibration

    Confirm resources, research coverage, sector guidance, independent appointments, pilot testing and the published edition rules.

  2. 6–4 months before publication

    2. Nominations and discovery

    Research and nominations enter one register with a common deadline.

  3. 4–3 months before publication

    3. Eligibility and evidence

    Verify roles and claims, complete candidate clarifications and establish assessable dossiers.

  4. 3–2 months before publication

    4. Independent assessment

    Complete valid ratings, moderation and integrity clearance.

  5. 2–1 months before publication

    5. Decisions and appeals

    Notify candidates, resolve appeals and rerun affected comparisons.

  6. Final month

    6. Publication clearance

    Recheck roles, confirm consent, clear reserves and complete process assurance.

  7. Publication and afterwards

    7. Recognition and corrections

    Publish the unranked edition, maintain its archive and monitor corrections.

Before nominations open, an edition must have complete eligibility rules, scoring and appeal procedures, independent decision-makers and alternates, confidential records, funded research and translation arrangements and a realistic public calendar.

Pilot testing covers varied hypothetical or permissioned dossiers. Sensitivity checks shift each weight by five percentage points and proportionally rescale the others to total 100, examining changes around the 70-point floor and selection boundary. Revisions are recorded before the cycle, not made mid-cycle to favour preferred candidates. Material live-cycle defects require independently documented remedies and fair treatment of everyone affected.

If a reliable cycle cannot finish before the Annual Summit, publication timing changes rather than standards. Any shortened schedule must preserve the same substantive safeguards.

14 / Selection policy

An inspectable decision trail

Each edition preserves its methodology version, approving authority, effective date, applicable edition and change summary. Accepted, amended and deferred changes are recorded separately; future amendments cannot become retrospective judgments of earlier honorees.

Records required by the policy
RecordWhat it preserves
Eligibility fileCandidate ID, verified role, enterprise type and boundary, operating countries, geographic finding, dates and all four executive tests with sources and reasons
Evidence registerPrecise claim, criterion, source provenance, period, units, scope, verification date and person, limitations and publication permission
Reviewer assessmentConflicts, scores or NA, source references, anchors, context, independent submission and attributable revisions
Moderation and committee recordEscalation triggers, third review, resolved facts, aggregation, ordering, ties, quorum, votes, recusals and reasons
Publication checklistEligibility and evidence gates, numeric floors, integrity clearance, appeals, roles, reserves, consent, equal recognition and process review, with owners and dates
Appeal recordNotification, permitted grounds, evidence, requested remedy, independent panel, determination and affected comparisons, preserving original and final decisions

Public corrections carry a dated change record. Confidential evidence and individual scores are not published to demonstrate transparency; transparency means explaining the rules, disclosing aggregate process information and retaining reproducible decisions for authorised review.

15 / Selection policy

References and honest limitations

A weighted qualitative assessment cannot provide an objectively complete or universally comparable measure of executive merit. Evidence availability, operating context, judgment and research coverage remain limitations. Calibration, multiple reviewers and records reduce avoidable inconsistency; they do not eliminate it.

The thresholds, weights, tenure and procedural periods are adopted programme choices, not externally certified best-practice constants. This policy is not jurisdiction-specific legal or data-protection advice.

External methodological references

Cartier Women’s Initiative: Regional Awards provides a reference for publishing eligibility, supporting evidence, assessment criteria and scoring guidance. Its constituency and ranked awards differ; its criteria and thresholds are not copied here.

OECD: Handbook on Constructing Composite Indicators (Statistics Working Papers 2005/03, DOI 10.1787/533411815016) informs the treatment of missing data, normalisation, weighting and aggregation. It does not validate this CEO methodology or its chosen weights.

Based on Selection Methodology, Document 03, prepared 13 October 2026; policy adoption confirmed by the initiative on 18 November 2026. Institutional attribution and Annual Summit details reflect subsequent confirmed instructions. Reference access dates recorded in the source document: 13 October 2026.

Questions or corrections

Accountability stays open.

Key questions

Who can qualify under the adopted policy?

Women with African citizenship who lead qualifying enterprises with substantive African operations must meet all four executive responsibility tests. At least 12 continuous months in the role are required; the assessment window is 36 months.

Ask a methodology question