Most owners and managers have opinions about who in their team performs well. Fewer have evidence. Without a fair way of measuring performance, decisions about pay, promotion, training and, occasionally, exit rest on impressions. Impressions favour the visible, the confident and the recently memorable over the consistently effective. Good people feel unrecognised, problems go unaddressed, and everyone suspects favouritism.
Measuring performance well is not about surveillance or catching people out. It aims to make expectations clear, give useful feedback, recognise good work, identify where coaching is needed, and make fairer decisions. This article covers why performance measurement matters, the main types of metrics, a range of practical methods, and how to turn measurement into conversations that actually improve performance.
Why measure performance
Done well, performance measurement helps a business:
- Assess performance objectively, both quantitatively (how much) and qualitatively (how well).
- Give constructive feedback based on facts rather than impressions.
- Track how the business is performing, since organisational results are the sum of individual and team results.
- Improve efficiency and productivity by revealing obstacles and good practices.
- Identify strengths and weak links in the team and in processes.
- Set fair pay and recognition.
- Balance workload by seeing who is overloaded and who has capacity.
- Evaluate creativity, accuracy and judgement, not just volume.
The purpose is improvement, not ranking people into winners and losers.
Types of performance metrics
Different roles need different measures. Broad categories include:
- Work quality: accuracy, defects, errors, rework and customer satisfaction.
- Work quantity: output volume, such as units, jobs, calls, sales or tickets resolved.
- Efficiency: output relative to time or cost, and turnaround times.
- Behaviour: reliability, collaboration, punctuality, adherence to policy and values.
- Skill and will: capability (skill) and motivation and commitment (will).
- Productivity: overall contribution relative to the resources used.
Most roles need a balance. Measuring only quantity encourages rushing. Measuring only behaviour can reward pleasantness over results.
Effort measures and result measures
A practical distinction is between effort scores, which are leading measures, and result scores, which are lagging measures.
- Effort scores measure the activities that drive results, such as sales calls made, quotes sent, inspections completed or preventive maintenance tasks done on schedule.
- Result scores measure the outcomes, such as revenue won, defects reduced or machine uptime.
For example, a salesperson’s expected effort might be 50 calls a day, and their expected result $100,000 in monthly sales. Someone making 40 calls has achieved 80% of the effort target. Someone making 60 has achieved 120%. The same applies to results.
Tracking both is powerful. High effort with low results suggests a skill, method or targeting problem, which calls for coaching and training. Low effort with good results may mean an unusually effective person or an easy territory. Low effort and low results raise questions of will, clarity or fit. Reviewing effort and results weekly lets managers intervene early.
Practical measurement methods
No single method is perfect. Most businesses combine several.
1. Self-evaluation
Ask employees to assess their own performance against their goals, highlighting achievements, difficulties and development needs. Self-evaluation encourages reflection and ownership, and comparing it with the manager’s view starts a useful conversation. The aim is to raise performance to meet expectations, not to lower expectations to match performance.
2. Checklists
For tasks with defined steps, a checklist both guides the work and provides a measure of completeness. A video shoot might have a pre-shoot checklist covering focus, framing, seating, background and sound. A machine setup might have a checklist of settings and checks. Reviewing completed checklists shows whether standards are followed and whether the same mistakes recur. In complex work, checklists are a help and are often essential for reliable results.
3. Performance and potential
Assess each person on two dimensions: performance, meaning results achieved so far, and potential, meaning capacity to grow into bigger responsibilities. A simple four-way grid follows:
| Low performance | High performance | |
|---|---|---|
| High potential | Coach: help them convert ability into results | Develop and stretch: future leaders |
| Low potential | Address: clarify expectations, support and fit | Value and develop in role: reliable contributors |
This helps target development effort. Be careful, though: “potential” is easily biased, so base it on evidence such as learning speed, judgement and initiative.
4. Customer feedback
Customers often see performance managers miss. Ask customers to rate their experience with an account manager, technician or service agent, for example on a 0–10 scale. High ratings are evidence of strong performance. Low ratings warrant investigation. This is related to, but not the same as, the Net Promoter Score, which asks how likely customers are to recommend the business and subtracts the percentage of detractors (0–6) from the percentage of promoters (9–10).
5. 180-degree feedback
Gather feedback from directly above and directly below the person, such as their manager and their direct reports. It is simple to start with a short form or template of questions. It also gives employees a chance to raise problems and comment on management.
6. 360-degree feedback
Gather feedback from all around the person: manager, peers, direct reports, other departments, customers and suppliers where relevant. It gives a rounded picture, especially for leadership and relationship-heavy roles. It requires care. Keep it anonymous where appropriate, use it for development rather than punishment, and have a skilled person help interpret the results.
7. Defects and errors
Where work produces tangible outputs, defect and error rates are objective quality measures. Corrections per edited video, drawing revisions caused by errors, rejected parts and software bugs are all examples. Compare like with like, since some work is inherently harder, and look for patterns that point to training or process issues rather than blame.
8. Behavioural rating scales
Rate specific behaviours that matter to your business on a defined scale. Examples include reliability, responsibility, collaboration, punctuality, accepting feedback, meeting deadlines, attention to detail, initiative and living the company’s values. Choose the five to eight behaviours that matter most, and describe what each looks like at different levels, so that ratings are consistent between managers.
9. Activity and output measures
For sales and service roles, typical measures include prospects contacted, calls made, meetings held, visits, active leads, first-contact resolution, average handling time and call quality. Customer relationship management systems can produce these daily. Always pair activity measures with result and quality measures, because activity alone is easily inflated.
10. Adherence to policies
Some standards are non-negotiable: safety rules, respectful behaviour, honesty and legal obligations. Make them explicit and apply them consistently. A zero-tolerance approach to serious misconduct such as harassment, discrimination or dishonesty protects both people and the business, provided it is backed by fair process.
11. Contribution to profit
Some businesses estimate the financial contribution of teams or roles by comparing the cost of a team with the value it generates. A department costing a known amount per year can be assessed on the revenue, margin or savings it delivers. This works best at team level, for revenue-linked roles, and needs reliable management accounts.
A note on forced ranking
Forced ranking requires managers to rank employees against each other, often with fixed proportions labelled top, middle and bottom. It is quick but crude. It can damage collaboration, encourage gaming and penalise strong teams, and many large organisations have moved away from it. If used at all, treat it as one input, not the decision.
Turning measurement into improvement
Measurement only helps if it leads to better conversations. Principles for effective reviews:
- Review frequently: monthly or quarterly check-ins, not just an annual appraisal. Problems addressed early are easier to fix.
- Document performance, both good and poor, with specific examples.
- Discuss the evidence openly and invite the employee’s view.
- Be future-focused: spend more time on plans for the next quarter than on past mistakes.
- Listen actively and ask solution-oriented questions: “What would help you hit this target?”
- Ask for feedback on your own management. It is a two-way process.
- Agree goals and development plans for the next period.
- Be transparent and collaborative: the employee should never be surprised by an assessment.
When performance does not improve after clear feedback, support and reasonable time, more formal steps may be needed. In Australia, fair process matters both ethically and legally. Document expectations, feedback and support, and seek HR or legal advice before any dismissal.
A simple framework for a small business
- Define three to five KPIs per role, mixing effort, result and quality measures.
- Choose two or three behaviours that reflect your values and are rated on a defined scale.
- Hold a monthly 30-minute check-in to review numbers, discuss obstacles and agree actions.
- Run a quarterly review using self-evaluation, manager assessment and, for some roles, customer or 180-degree feedback.
- Hold an annual development conversation about career goals, potential and training plans.
Common biases to guard against
Even with good metrics, human judgement creeps into every assessment. Knowing the common biases helps managers correct for them:
- Halo and horns effects. One strong trait, good or bad, colours the whole assessment. A charismatic salesperson is rated highly on administration they neglect, and a quiet technician’s excellent quality goes unnoticed.
- Recency bias. The last few weeks dominate the memory of a whole quarter or year. Keeping brief notes throughout the period counters this.
- Leniency and central tendency. Some managers rate everyone highly to avoid conflict. Others rate everyone as average to avoid commitment. Either way, the ratings stop carrying information.
- Similarity bias. People rate those who resemble them, in background, personality or working style, more favourably.
- Attribution errors. Managers tend to blame individuals for poor results that were caused by the process, the equipment or the workload, and to credit individuals for results that were really driven by circumstances.
- Visibility bias. People who work near the manager, or who talk about their work, get noticed more than those who quietly deliver.
Practical counters include written expectations at the start of each period, notes kept throughout, evidence-based ratings with examples, calibration discussions where managers compare ratings across teams, and asking “what else could explain this result?” before concluding that a person is the cause.
A worked example: a small fabrication shop
Consider a fabrication business with twelve staff: four fabricators, two welders, a setter, a storeperson, an estimator, a drafter, an office manager and a supervisor. Until now, pay reviews have been based on the owner’s impression and length of service. Two good fabricators have recently left, saying their efforts went unrecognised.
The owner introduces a simple system:
- For fabricators and welders, three KPIs: jobs completed against standard hours (efficiency), first-pass inspection rate (quality) and safety observations (behaviour). Data comes from job cards and inspection records the supervisor already keeps.
- For the estimator, quotes issued per week (effort), quote turnaround time (efficiency), win rate (result) and quoting errors found later (quality).
- For the drafter, drawings released per week, revisions caused by drafting errors and on-time release against the production schedule.
- For everyone, three behaviours rated on a four-point scale with written descriptions: reliability, teamwork and following safe work procedures.
Each person has a monthly fifteen-minute check-in with the supervisor or owner to look at the numbers and discuss obstacles. Each quarter, a longer review combines a short self-evaluation, the KPIs, behaviour ratings and the manager’s comments. Pay decisions at the annual review draw on the four quarterly reviews.
Within two quarters, the owner learns several things impressions had hidden. One welder with an average reputation had the best first-pass quality in the shop. The estimator’s low win rate was traced to slow turnaround, because quotes sat waiting for drawings, which pointed to a process problem rather than a personal one. And two fabricators who seemed slow were consistently given the most complex jobs. The system did not just rate people. It improved how work was allocated and supported.
Frequently asked questions
How many metrics should one person have? Usually three to five. More than that dilutes attention and makes reviews mechanical.
Should metrics be linked to bonuses? They can be, but be cautious. Strong financial incentives on narrow metrics encourage gaming. Many small businesses use metrics mainly for feedback and development, and link pay to a broader overall assessment.
What about roles that are hard to measure? Use a combination of outputs, such as deliverables completed and deadlines met, internal customer feedback and behaviour ratings. Agree upfront what good looks like, even if it cannot be counted precisely.
What if an employee disagrees with their assessment? Listen, review the evidence together and correct anything that is wrong. If disagreement remains, record both views and agree specific expectations for the next period. A fair process matters more than winning the argument.
Is annual appraisal enough? Rarely. By the time an annual review arrives, feedback is stale. Short, regular check-ins with a fuller quarterly or six-monthly review work better.
Summary
Fair performance measurement replaces impressions with evidence and turns feedback into improvement. Use a balance of quality, quantity, efficiency and behaviour measures, and track both effort and results. Combine methods such as self-evaluation, checklists, customer feedback, 180- or 360-degree feedback, defect rates and behavioural rating scales, while treating crude tools like forced ranking with caution. Above all, review often, discuss openly, focus on the future and remember that the goal is to help people improve.
Sources: small-business training notes on employee performance metrics and business performance frameworks, together with common human-resources practice. This article is general information, not legal or HR advice.
