Workflow 7
Performance Management
Continuous feedback with an annual spine
Illustrative current-practice design
This workflow shows a conventional enterprise performance-management cycle: annual goals and ratings supported by continuous conversations, mid-year review, calibration and differentiated rewards.
Use it as a complete starting point, then adjust it for your company's strategy, culture, workforce, jurisdiction, collective agreements, technology, risk appetite and future-of-work direction.
Content last reviewed 28 July 2026.
Philosophy
These conversations belong between managers and employees. HR provides the framework and protects the fairness, and nothing in the review should be a surprise.
Cadence
Goals in January, mid-year in July, annual cycle November to January
The steps
- 01
January: goal setting
- Manager and employee set goals together
- Goals logged in HRIS
- 02
Continuous
- Weekly one-to-ones with every direct report are the baseline in this example configuration. Companies may choose a different reliable cadence based on work type, team size, employee needs and manager capacity.
- Feedback shared informally through HRIS or Teams throughout the year
- Wins recognised publicly in team meetings
- 03
July: mid-year
- Employee self-assesses goal progress and support needs
- Manager discussion; action items logged in HRIS
- HR involved if performance concerns emerge
- 04
November: annual review
- Employee self-assessment opens the cycle
- In this example configuration, 360 feedback is optional for individual contributors and required for managers. The employee nominates three to five contributors and the HRBP reviews the panel for balance. Responses are confidential and synthesized rather than attributed.
- Manager rates goal achievement and competencies on a five-point scale
- 05
December: calibration
- HRBP facilitates a calibration session with all managers in the business unit
- Ratings reviewed together for consistency against a reference distribution
- 06
January: delivery and merit
Legally sensitive: verify locally- Manager delivers feedback using the SBI model, communicates the rating clearly, and names development priorities
- Merit differentiated by rating. Top performance: 4–6%. Strong performance: differentiated increases within the available pool. Solid performance: increase aligned with the approved company matrix. Below expectations: reduced or no merit increase, subject to policy and local requirements.
- Lower ratings trigger a review of whether structured performance support or a formal PIP is appropriate. The rating alone does not determine the response. Consider the evidence, previous feedback, clarity of expectations, manager support, role design, system barriers and any accommodation or protected-leave considerations.
Who does what (RACI)
| Step | Responsible | Accountable | Consulted | Informed |
|---|---|---|---|---|
| Goal setting | Employee | Manager | HRBP | HR Ops |
| 360 panel approval | HRBP | HRBP | Manager | Employee |
| Calibration | Managers | BU Leader | HRBP | CHRO |
| Merit decision | Manager | BU Leader | Rewards COE, Finance | Employee |
Stress tests
Reviews not completed
System reminders, escalation to leadership, pay increases held.
Ratings cluster high
Calibration surfaces the distribution and requires justification from leadership.
Employee disputes a rating
HRBP mediates between employee and manager before any escalation.
New managers rating inconsistently
Coaching plus pairing with an experienced peer during calibration.
Design notes
The five-point rating scale
Five-point scales remain common because they create differentiation for talent, performance and compensation decisions. Alternatives include fewer rating categories, narrative assessments, goal-based outcomes or rating-free development conversations. The organization should choose deliberately and apply the approach consistently.
Continuous feedback signals
Supplement periodic 360 feedback with more continuous, consent-aware signals from project teams, internal clients and cross-functional collaborators. Do not use AI-generated performance conclusions without human review, transparent criteria and appropriate privacy and bias controls.
Calibration and the reference distribution
Calibration compares evidence and tests whether ratings are being applied consistently. A reference distribution may be used to identify unexplained concentration or inconsistency.
The merit percentages
These percentages illustrate how differentiated merit may operate. They are not market recommendations. Recalibrate them each cycle for the approved budget, inflation, market movement, internal equity, range position, pay-transparency obligations and the company's compensation philosophy.
Future-of-work considerations
- Goals updated as business priorities change
- Project- and team-based evidence
- Skills and capability growth alongside business outcomes
- Recognition of collaboration and knowledge contribution
- Shorter check-ins supported by technology
- Reduced dependence on a single manager's judgment
- AI-assisted administration without AI-determined ratings
- Clear boundaries on employee monitoring and behavioural data
- Manager capacity as an explicit design constraint