Introduction
Retail trade is one of Australia’s largest employing industries and one of the hardest in which to manage performance. Work is dispersed across hundreds of sites, much of the workforce is part-time or casual, and store managers supervise teams whose rostered hours rarely overlap (Australian Bureau of Statistics [ABS], 2024). A system built for a salaried employee sitting near a manager will not survive a shop floor in Geelong on a Thursday night.
This coursework designs a performance management system for Harbourline Retail Group, a hypothetical homewares chain of approximately 900 staff across 46 stores in New South Wales, Victoria and Queensland. Of these, 210 are salaried and 690 are award-covered team members, mostly part-time or casual under the General Retail Industry Award 2020. The chain currently runs one appraisal each June, on a paper form, with a five-point rating feeding a discretionary pay review.
Five tasks follow: a critique of annual appraisal against continuous performance management; a competency framework and weighted balanced scorecard; calibration and rating distribution; remuneration and underperformance management under the Fair Work Act 2009 (Cth); and implementation risk.
Task 1: Annual Appraisal Against Continuous Performance Management
The limitations of the annual event
DeNisi and Murphy (2017), reviewing a century of research, concluded that measurement has improved but that appraisal does not reliably improve performance, which is the outcome organisations actually purchase. Ratings assigned once a year attract recency bias, are contaminated by halo effects, and are distorted by the rater’s awareness that the rating drives pay, which produces leniency (Murphy, 2020).
The second limitation is temporal. Feedback delivered eleven months after the behaviour it describes cannot change that behaviour. Cappelli and Tavis (2016) documented employers replacing annual ratings with frequent forward-looking conversations, arguing that the annual cycle suits stable hierarchical work rather than environments where priorities reset quarterly. In retail the mismatch is severe: trading conditions, staffing and range change several times within one appraisal period.
The third limitation is administrative. Aguinis (2019) distinguishes appraisal, the act of evaluating, from performance management, a continuous process of developing performance in alignment with strategy. Conflated, the system becomes a compliance exercise in which the completion rate is the metric.
The limitations of abolishing ratings
The opposite error is equally costly. Pulakos et al. (2019) caution that organisations which removed ratings still had to decide pay, promotion and termination, and made those decisions less transparently on undocumented impressions; evaluation goes underground rather than disappearing (Murphy, 2020). This matters legally in Australia, because dismissal for unsatisfactory performance requires warnings and documented evidence that a purely conversational system does not generate.
The design adopted here is therefore hybrid. Continuous coaching carries the developmental load through monthly check-ins, while one calibrated evaluation each year carries the administrative load of pay and progression. This mirrors Australian practice, where organisations have shortened cycles and added check-ins rather than abolishing ratings (Australian HR Institute [AHRI], 2023).
Task 2: System Architecture
Figure 1 sets out the proposed cycle. Four of its five stages run continuously and only Review is annual; each year’s development plan becomes an input to the next year’s goals.
The competency framework
Goals describe outcomes without describing the conduct that produces them: a store that hits budget through aggressive selling that generates complaints has not performed well. Table 1 therefore defines five competencies with observable behavioural indicators and a four-level proficiency scale, so that one language runs from a casual in Parramatta to an assistant manager in Brisbane, with only the expected level differing.
Table 1: Harbourline competency framework, showing behavioural indicators and expected proficiency by role
| Competency | Observable behavioural indicators | Proficiency anchors | Expected: team member | Expected: store manager |
|---|---|---|---|---|
| Customer engagement | Acknowledges customers within 30 seconds; asks needs-based questions before recommending; applies consumer guarantee obligations accurately | L1 Developing: applies with direction. L2 Capable: independent in routine situations. L3 Proficient: independent in complex situations, explains reasoning. L4 Leading: sets the standard, coaches others | L2 | L4 |
| Commercial judgement | Reads the weekly trade report; adjusts rosters to traffic; protects margin through markdown discipline | As above (trading decisions) | L1 | L4 |
| Team leadership and coaching | Runs pre-shift briefings; gives specific feedback within 48 hours; documents a monthly check-in for every direct report; distributes desirable shifts transparently | As above (people leadership) | Not assessed | L4 |
| Safety and compliance | Completes hazard inspections; corrects manual handling and trip risks; de-escalates and reports aggressive customer incidents | As above (risk) | L2 | L3 |
| Operational execution | Achieves planogram compliance at range change; maintains stocktake accuracy; fulfils online orders within service windows | As above (process) | L2 | L3 |
Expected levels differ by role, not by individual, which stops a supervisor’s personal expectations becoming the standard. Team leadership is not assessed for team members, because rating people against competencies they cannot demonstrate produces noise that contaminates pay decisions.
The balanced scorecard and a worked rating
Outcome measures use the four perspectives of the balanced scorecard, developed to stop financial measurement crowding out the drivers of future performance (Kaplan & Norton, 1996). Weightings sum to 100 per cent and are set centrally, so that a manager in a low-growth regional catchment is not disadvantaged against a metropolitan one. Targets are difficult but attainable, consistent with evidence that specific challenging goals outperform vague exhortation (Locke & Latham, 2019). Table 2 applies the scorecard to the manager of Store 214, Geelong.
Table 2: Balanced-scorecard rating for a sample store manager, Store 214, financial year 2025
| Perspective | Measure | Target | Weighting | Actual | Score (1-5) | Weighted contribution |
|---|---|---|---|---|---|---|
| Financial | Store sales against budget | 100% | 20% | 104% | 4 | 0.80 |
| Financial | Gross margin percentage | 34.0% | 10% | 33.2% | 3 | 0.30 |
| Customer | Net promoter score | +40 | 15% | +47 | 5 | 0.75 |
| Customer | Mystery shop assessment | 85% | 10% | 82% | 3 | 0.30 |
| Internal process | Stock loss as a share of sales | 1.20% or lower | 10% | 1.05% | 4 | 0.40 |
| Internal process | Safety and compliance actions closed on time | 100% | 10% | 100% | 4 | 0.40 |
| Learning and growth | Competency assessment against Table 1 | L3 Proficient | 15% | L4 Leading | 4 | 0.60 |
| Learning and growth | Team learning completion | 95% | 10% | 88% | 2 | 0.20 |
| Total | 100% | 3.75 |
The overall result is the sum of each weighting multiplied by its score:
Weighted score = (0.20 x 4) + (0.10 x 3) + (0.15 x 5) + (0.10 x 3) + (0.10 x 4) + (0.10 x 4) + (0.15 x 4) + (0.10 x 2)
Weighted score = 0.80 + 0.30 + 0.75 + 0.30 + 0.40 + 0.40 + 0.60 + 0.20 = 3.75 out of 5.00
The arithmetic is less interesting than what it exposes. This manager sits well above expectation on customer advocacy and sales and well below it on team learning, which the composite 3.75 conceals. The review is therefore held against the rows of Table 2 rather than the total, and the development plan attaches to the weakest row.
Task 3: Calibration and Rating Distribution
Weighted scores are computed by the manager, then moderated by a calibration panel of the state manager, the eight district managers and a human resources representative, who review every salaried employee’s evidence before release. Calibration addresses measurement error rather than malice. In the pilot, the proportion rated Exceeds or Outstanding ranged from 61 per cent in one district to 18 per cent in another, while their sales-to-budget results differed by only 1.8 percentage points. After calibration the range narrowed to 22-38 per cent.
Rating inflation is also a financial control issue. The 210 salaried employees earn an average base salary of A$78,000, a base cost of 210 x A$78,000 = A$16,380,000. One percentage point of drift in the average merit increase therefore costs 0.01 x A$16,380,000 = A$163,800 a year, and because base salary compounds, that cost recurs.
The system publishes an indicative distribution (Table 3) but enforces no quota. Forced distribution applied to a store team of nine produces the absurd requirement that one person be rated Below expectations regardless of performance, and quotas invite cynicism and gaming (Pulakos et al., 2019). Panels must instead justify departures from the indicative shape with evidence.
Calibration also carries an equity function. Ratings are audited annually by gender, age band and employment type, because part-time employees, disproportionately women, are vulnerable to being rated on visibility rather than output. Outcomes correlated with a protected attribute expose the employer to general protections claims under the Act. Transparency also drives perceived fairness, which predicts acceptance even of unfavourable outcomes (Colquitt & Rodell, 2015), so panel membership, decision rules and appeal rights are published in advance.
Task 4: Remuneration and Underperformance Management
Remuneration outcomes
Table 3 translates weighted scores into pay outcomes for salaried employees. Award-covered team members receive award rates and any annual wage review increase irrespective of rating; their ratings drive classification progression, development access and a store team bonus, not individual base pay. This separation prevents the system from appearing to displace award entitlements.
Table 3: Rating bands, indicative distribution and remuneration outcomes for salaried employees
| Weighted score | Rating | Indicative distribution | Base pay movement | Incentive multiplier | Other consequence |
|---|---|---|---|---|---|
| Below 2.00 | Below expectations | 5% | 0% | 0.00 | Formal performance improvement plan |
| 2.00 to 2.74 | Partially meets | 10% | 0 to 1.0% | 0.00 | Documented improvement objectives |
| 2.75 to 3.49 | Meets expectations | 50% | 2.5% | 0.92 to 1.16 | Standard development plan |
| 3.50 to 4.24 | Exceeds expectations | 25% | 3.5% | 1.17 to 1.41 | Considered for stretch assignment |
| 4.25 to 5.00 | Outstanding | 10% | 4.5% | 1.42 to 1.50 (capped) | Entry to the succession pool |
The multiplier is derived from the weighted score rather than assigned by band, so outcomes vary continuously and managers cannot win a step change by arguing a candidate over a boundary. For the sample manager, on a base salary of A$96,000 with a 12 per cent incentive target:
- Target incentive = 0.12 x A$96,000 = A$11,520
- Payout factor = weighted score / 3.00 = 3.75 / 3.00 = 1.25
- Incentive payable = A$11,520 x 1.25 = A$14,400
- Base pay movement (Exceeds band) = 0.035 x A$96,000 = A$3,360, giving a new base of A$99,360
Two gates apply before any incentive is paid: gross margin of at least 32.0 per cent, and no critical safety action outstanding at year end. Store 214 recorded 33.2 per cent and closed all actions, so both are satisfied. Gates matter because a purely additive scorecard lets a strong sales result purchase tolerance for a safety failure.
Underperformance and procedural fairness
Where a rating falls below 2.00 the system moves from development to formal management, and is then constrained by law rather than preference. Harbourline employs 900 people, so the Small Business Fair Dismissal Code does not apply; employees past the six month minimum employment period may apply to the Fair Work Commission. In deciding whether a dismissal was harsh, unjust or unreasonable, the Commission must consider the section 387 criteria of the Fair Work Act 2009 (Cth): a valid reason related to capacity or conduct, notification of that reason, an opportunity to respond, access to a support person, and prior warnings about unsatisfactory performance.
The performance improvement plan is engineered to generate that evidence. It runs for six weeks and specifies the standard, the gap, measurable objectives, the support provided (coaching, refresher training, adjusted rosters where personal circumstances contribute) and the consequence of non-improvement. Meetings are fortnightly, the right to a support person is advised in writing, and outcomes are documented and given to the employee (Fair Work Ombudsman, 2024). Managers are trained to separate performance, a capacity matter warranting support and time, from misconduct, which follows an investigative pathway. Because most team members are covered by the General Retail Industry Award 2020, meetings sit within rostered and paid hours.
Task 5: Implementation Risks and Change Management
Australian evidence suggests that redesigns fail at the manager capability layer rather than the system layer (AHRI, 2023). The principal threats are:
- Manager capability. Monthly coaching requires skills that many retail managers, promoted for sales results, were never taught. Mitigation: a two-day workshop before launch, then observed practice conversations assessed against the Team leadership competency.
- Time cost. Twelve check-ins for eight direct reports is a real addition to a manager’s week. Mitigation: 20-minute check-ins on a three-question template, scheduled inside rostered hours.
- Rostering mismatch. A casual working two shifts a week cannot be managed on a salaried cadence. Mitigation: a quarterly cycle for casuals, delivered by the supervisor on shift.
- Data quality. Net promoter and mystery shop data must be reliable at store level or the scorecard loses legitimacy. Mitigation: a minimum sample threshold per store, below which the weighting is redistributed.
- Consultative risk. The award’s consultation term requires consultation about major workplace change with significant effects on employees. Mitigation: employee representatives are consulted during design, not after sign-off.
The change approach follows Kotter’s (2012) sequence over a 12 month runway. Urgency comes from presenting the district rating variance and the A$163,800 cost of one point of drift, since a demonstrated inconsistency in managers’ own results moves more people than an abstract appeal to fairness. A guiding coalition of the retail operations director, three respected store managers and the human resources manager owns the design, answering the credibility problem that arises when a system is seen as a support office imposition. The change anchors when check-in completion and calibration participation enter the district managers’ own scorecards.
Conclusion
The annual appraisal at Harbourline fails on three grounds: it measures late, it measures with bias, and it does little to improve what it measures (DeNisi & Murphy, 2017; Murphy, 2020). Abolishing ratings, however, would remove the evidence Australian employers need for defensible pay and dismissal decisions. The hybrid proposed here keeps one calibrated annual rating, shifts developmental weight onto monthly coaching, and measures behaviour and outcomes through a competency framework and a transparent weighted scorecard. The worked example returned 3.75, but its more important output was the weakness in team learning that a single rating would have hidden. Calibration panels, an indicative rather than forced distribution, and an improvement pathway built on the section 387 criteria make the system defensible to an employee and to the Fair Work Commission. Its remaining condition of success is manager capability, which is why implementation invests more in coaching skills than in the system that records them.
References
Aguinis, H. (2019). Performance management (4th ed.). Chicago Business Press.
Australian Bureau of Statistics. (2024). Labour force, Australia, detailed. ABS.
Australian HR Institute. (2023). Performance management practices in Australian organisations. AHRI.
Cappelli, P., & Tavis, A. (2016). The performance management revolution. Harvard Business Review, 94(10), 58-67.
Colquitt, J. A., & Rodell, J. B. (2015). Measuring justice and fairness. In R. S. Cropanzano & M. L. Ambrose (Eds.), The Oxford handbook of justice in the workplace (pp. 187-202). Oxford University Press.
DeNisi, A. S., & Murphy, K. R. (2017). Performance appraisal and performance management: 100 years of progress? Journal of Applied Psychology, 102(3), 421-433.
Fair Work Act 2009 (Cth).
Fair Work Ombudsman. (2024). Managing performance and warnings. Australian Government.
General Retail Industry Award 2020 (MA000004).
Kaplan, R. S., & Norton, D. P. (1996). The balanced scorecard: Translating strategy into action. Harvard Business School Press.
Kotter, J. P. (2012). Leading change. Harvard Business Review Press.
Locke, E. A., & Latham, G. P. (2019). The development of goal setting theory: A half century retrospective. Motivation Science, 5(2), 93-105.
Murphy, K. R. (2020). Performance evaluation will not die, but it should. Human Resource Management Journal, 30(1), 13-31.
Pulakos, E. D., Mueller-Hanson, R., & Arad, S. (2019). The evolution of performance management: Searching for value. Annual Review of Organizational Psychology and Organizational Behavior, 6, 249-271.