Choose between common models: the Welch/GE 20-70-10 (top 20% rewarded, vital 70% developed, bottom 10% managed out), a five-tier model (10-20-40-20-10), or a three-tier model (top, middle, bottom). Consider organizational size, culture, and maturity — forced ranking works differently in a 50-person startup versus a 50,000-person enterprise. Document the chosen model with clear rationale.

Create detailed descriptors for each rating category — for example, 'Exceptional' (top tier): consistently exceeds all objectives, demonstrates leadership behaviors that elevate team performance, delivers breakthrough results. 'Below Expectations' (bottom tier): falls significantly short on key objectives despite adequate support, does not meet role requirements. Specificity reduces subjectivity in calibration.

Forced distribution becomes statistically questionable with small teams. Set a minimum threshold — typically 25–30 employees per calibration group — below which the distribution is treated as a guideline rather than a mandate. For smaller teams, consider pooling into larger calibration groups across departments.

Decide whether the distribution curve is mandatory (managers must place exactly the specified percentages in each category) or advisory (managers should approximate the distribution but can deviate with documented justification). Hard mandates create consistency but can force unfair ratings; soft guidelines allow flexibility but risk grade inflation.

Acknowledge the documented risks of forced ranking: potential for discrimination claims if protected groups are disproportionately placed in lower tiers, negative impact on team collaboration, and reduced psychological safety. Microsoft famously abandoned stack ranking in 2013 after attributing innovation declines to its competitive dynamics. Implement safeguards such as bias audits, appeal processes, and regular program review.

Provide each manager with a comprehensive data package for their team members including goal achievement scores, competency ratings, project contributions, peer feedback, and any other relevant performance data. Managers should prepare a preliminary rating for each employee with supporting evidence before the calibration meeting.

Convene managers from related teams or departments in a facilitated session where each presents their talent and proposed ratings. Cross-functional calibration ensures that standards are consistent — a 'high performer' in one team meets the same standard as in another. The HR facilitator challenges inconsistencies and ensures evidence-based discussion.

Begin by discussing employees proposed for the top tier and bottom tier first, as these are the most consequential placements. Use evidence-based debate to reach consensus. Then distribute the remaining employees across middle tiers. Allow the conversation to adjust initial placements as comparative discussion reveals relative performance levels.

Record why each employee was placed in their tier, referencing specific data points and calibration discussion outcomes. This documentation is essential for manager-employee conversations, potential appeals, and legal defensibility. It also provides valuable data for assessing calibration quality over time.

Before locking ratings, analyse the distribution by gender, ethnicity, tenure, age, and other demographics to identify potential bias patterns. If any group is statistically over-represented in lower tiers, investigate whether the ratings are justified by performance data or whether bias may be influencing outcomes.

Equip managers with conversation frameworks for each tier. Top-tier conversations should reinforce specific behaviors to continue and discuss stretch opportunities. Bottom-tier conversations must be delivered with empathy, focus on specific performance evidence, and clearly outline expectations and support for improvement. Role-play these conversations in training sessions.

Top-rated employees should receive tangible recognition — accelerated salary increases, larger bonuses, stock grants, promotion fast-tracking, or high-visibility assignments. If the rewards for top performance are not meaningfully different from middle-tier rewards, the forced distribution loses its motivational purpose.

The middle tiers (representing 60–70% of employees) are the organizational backbone. Avoid treating them as 'average' — instead, provide targeted development plans, skill-building opportunities, and clear guidance on what they need to demonstrate for top-tier placement next cycle. Engagement of this group is critical to organizational success.

Bottom-tier employees should receive direct feedback, clear performance expectations, and a structured improvement plan (similar to a PIP) with a defined timeline. Provide genuine support for improvement while being transparent about the consequences of sustained underperformance.

Create a transparent mechanism for employees to challenge their rating through a review panel comprising senior HR and uninvolved managers. The appeal should consider whether the correct process was followed, whether evidence supports the rating, and whether any procedural unfairness occurred. Document and communicate the appeals process to all employees.

Track engagement survey scores, voluntary attrition rates, and internal mobility data segmented by rating tier. Research by the Institute for Corporate Productivity found that forced ranking can reduce collaboration and increase unhealthy competition. If engagement declines, consider softening the distribution or introducing mitigating practices.

Audit outcomes for bottom-tier employees: are they receiving genuine development support, or is the bottom tier simply a precursor to termination? If the overwhelming majority of bottom-tier employees exit, the system may be functioning as a de facto termination mechanism rather than a performance improvement tool.

Assess whether forced ranking is encouraging or inhibiting collaboration. In environments where teamwork is essential, stack ranking can create perverse incentives to undermine colleagues. If this dynamic is detected, consider team-based performance components or adjusting the distribution model.

The appropriate distribution percentages may change as the organization matures, market conditions shift, or talent strategy evolves. An organization that has invested heavily in hiring and development may genuinely have fewer low performers, making a rigid bottom-tier quota counterproductive. Adjust the model based on data.

Explore models that differentiate performance without mandating fixed percentages — such as calibrated assessments with recommended but flexible distributions, or relative performance rankings without forced bottom-tier quotas. Companies like Adobe and Deloitte have moved to check-in models that maintain accountability without the rigid curve.

Form a cross-functional committee including senior HR, legal, business leaders, and employee representatives to govern the program. The committee reviews aggregate data, hears appeals, assesses program fairness, and recommends adjustments. This governance structure provides accountability and protects against misuse.

Perform rigorous statistical analysis (chi-square tests, regression analysis controlling for performance variables) to determine whether rating distributions are equitable across demographic groups. Publish aggregate findings to leadership and take corrective action where disparities cannot be explained by legitimate performance differences.

Survey the competitive landscape to understand whether peer organizations use forced ranking, what distribution models they employ, and whether the trend is toward or away from this approach. As of recent years, many technology and professional services firms have moved away from forced ranking, while some manufacturing and financial services firms retain it.

Run an annual confidential survey assessing perceptions of fairness, accuracy, impact on motivation, and suggestions for improvement. Pay particular attention to feedback from middle-tier employees, who research shows are often the most demoralised by forced ranking systems because they feel overlooked.

Present the governance committee with a comprehensive annual review covering program outcomes, engagement impact, legal exposure, manager feedback, and industry trends. Make an evidence-based decision about whether the program continues to serve the organization's strategic talent objectives or whether an alternative approach would be more effective.