Performance Management and Appraisal

An operator at a Pune auto-parts plant hears at his April review that his rejects were too high all last year. Nobody told him in June, when he could still have fixed it. Performance management exists so that the review is never the first time he hears.

Performance appraisal is one event inside that routine: a formal, periodic evaluation of how well a person did in the period just ended, usually once or twice a year.

Objectives of performance management. There are four:

• Goal alignment: directing each person's effort towards the organisation's strategic outcomes.
• Continuous improvement: building a high-performance culture through regular feedback and coaching.
• Employee development: finding skill gaps, training needs and growth opportunities.
• Recognition and motivation: raising morale through rewards, bonuses and recognition, such as an Employee of the Month award.

The five-phase cycle. The phases repeat, and what the last one learns feeds the next round of planning.

Four key terms run through the cycle.

The gap, worked. The fitter makes 5,000 parts. The standard allows 2%, which is 100 rejects. He has 5%, which is 250 rejects. The gap is 250 − 100 = 150 parts, or 3 percentage points. A good plan tackles why: the set-up, the tool wear, or his skill. Measures like KPIs are often grouped under four headings (financial, customer, internal process, learning and growth) in the Balanced Scorecard, which has its own chapter.

Don't confuse performance management with performance appraisal. Appraisal is the formal look back; management is the whole year of planning, watching, helping and rewarding. Textbooks also name the five phases as planning, monitoring, developing, rating and rewarding. It is the same loop with different labels.

Where it breaks. On paper the cycle is tidy; in practice, managers often skip the middle phases, and it shrinks to a form filled in once a year. A KPI also tends to steer behaviour: if only rejects are counted, speed or teamwork may suffer.

In practice. Standards come from the job description, which comes from job analysis. The gap tells the manager whether the fix is training, tools or something else.

For exams. Know the contrast in the first table, the five phases in order, and the four key terms. SMART is a favourite.

Plan, monitor, feedback, review, reward: then plan again.

A bank branch manager must name the best of twelve clerks by Friday. If she lines them up from best to worst, she gets one answer; if she scores each one on punctuality and teamwork, she may get another. The method decides what she sees.

Paired comparison, counted. With N employees, the number of pairs is N × (N − 1) ÷ 2. For a team of six clerks it is 6 × 5 ÷ 2 = 15 comparisons. For all twelve it is 12 × 11 ÷ 2 = 66. The clerk who wins the most pairs ranks first.

Confidential report. In its classic form, as in the Annual Confidential Report of government service, the employee does not see the report. Feedback comes only if there is an adverse entry. Ratings are tied to promotion, so they are open to manipulation. Many government services have moved towards open reports that the officer sees.

Other traditional methods. Textbooks list several more:

• Forced distribution: the rater must spread employees across the scale in fixed shares, on the assumption that performance follows a bell curve. It stops everyone being rated high, but it is unfair to a strong team.
• Forced choice: the rater picks, from sets of statements, the one that fits best. Which statements count for more is hidden from the rater, so favouritism has less room.
• Essay method: the rater writes a free description of strengths, weaknesses and promotion potential. It fills gaps a checklist leaves, but depends on the rater's writing and memory.
• Field review: someone from outside the department, often from HR, reviews the employee. It gives comparable ratings for promotion, but the reviewer does not see the daily work.
• Performance tests and observation: a test of knowledge or skill, written or practical. Tests must be reliable and valid, and can be costly to build.

Don't confuse ranking with rating. Ranking compares people with each other, so someone must come last even in a strong team. Rating compares each person with a scale or standard, so everyone can score high. Judging people against others rather than a standard is the contrast effect, a perception error.

Where it breaks. Most of these methods rate traits, not results, and look back rather than forward. They leave room for personal bias, and several give the employee little to act on. That is why firms added the modern methods.

For exams. Match each description to its name. Remember that the critical incident technique is also a job-analysis tool, and that paired comparison has its own formula.

Rank, pair, scale, tick, incident, secret report.

A bank wants to know more than what a branch officer achieved. It wants to know how she treats juniors, and whether she could run a branch. One manager's yearly rating cannot answer both.

360-degree feedback. Anyone with useful information about how a person works can rate them.

Management by objectives (MBO) was set out by Peter Drucker in 1954. The process has four steps:

1. Set goals and desired results for each subordinate, jointly.
2. Set performance standards.
3. Compare actual results with the goals.
4. Set new goals, and new plans for the goals that were missed.

A sales officer might agree to bring in forty new accounts in the year. At the review, the count of accounts opened, not the manager's impression, decides the rating.

BARS was developed by Smith and Kendall in 1963. Its builders gather examples of effective and ineffective behaviour for a job and place them on the scale as anchors. Here is a customer-handling scale for a bank teller (an illustration):

The rater picks the description that matches what they saw, which is harder to bend than choosing "good" or "average".

Assessment centres began in military officer selection, before and during the Second World War. Candidates work through in-basket exercises, group discussions, role plays and simulations, while trained observers score them against set criteria such as planning, decision making and persuasion. They suit promotion to management but cost a lot.

Psychological appraisal aims at future potential. It uses in-depth interviews, psychological tests and talks with supervisors to look at emotional, intellectual and motivational traits.

Human resource accounting sets the cost of recruiting, training and keeping an employee against the value they generate. In textbooks it is also called the cost accounting method of appraisal.

Don't confuse BARS with a graphic rating scale. A graphic scale labels its points with words such as poor, average and good, applied to traits. BARS labels each point with a description of behaviour someone can observe.

Where it breaks. No modern method removes rater bias; BARS still suffers the distortions common to rating, and a 360 can turn into a popularity contest. MBO can push people to hit easy numbers and neglect what is not measured. Many of these methods are also costly, so small firms rarely use them all.

For exams. Match names to ideas: Drucker with MBO, Smith and Kendall with BARS, simulations with assessment centres, "potential" with psychological appraisal.

Go deeper for practitioners

A decision. Your plant adds 360-degree ratings for supervisors and links them to the annual bonus. Ratings from subordinates rise sharply, and several supervisors begin trading favourable marks. Do you keep the pay link or drop it? When pay depends on ratings, raters may inflate or trade marks. Many firms keep 360 for development and use other evidence for pay.

Links. The cross-cultural chapter (this unit) warns that where power distance is high, juniors may rate every boss at the top; anonymity helps. The Johari window in interpersonal behaviour (this unit) explains the value of 360: the blind area is what others see in you and you do not.

360 sources, objectives, behaviour anchors, simulations, psychologists, money.

Two supervisors rate the same fitter. One gives him 9 out of 10 on everything, the other 5 out of 10 on everything. Neither looked at his work record. The problem lies in the raters, not in the fitter.

Leniency and strictness work in opposite directions. Both, with central tendency, make a system useless, because if everyone gets the same rating it tells the firm nothing about who differs.

Don't confuse the halo effect with an appraisal method. The halo effect is a rater error, in which one trait colours the whole judgement. BARS, MBO and 360-degree feedback are methods used to appraise. An exam question that asks which of these is not a method of performance appraisal has the halo effect as its answer. The perception chapter (this unit) covers the halo and horn effects as general perception errors.

Making appraisal fair. No rule removes these errors, but a few habits cut them:

• Use clear standards drawn from the job description, not general impressions.
• Rate one dimension for all employees before moving to the next, to weaken the halo.
• Keep a running record of incidents all year, against recency and first impression.
• Use more than one rater, or hold a calibration meeting where supervisors compare ratings.
• Train raters, and let the employee comment on or question the rating.

Why firms bother: benefits and challenges.

Appraisal also feeds decisions: who is promoted, whose pay rises, who needs training, and who is not meeting standards.

The feedback conversation. A review is only useful if the person leaves knowing what to do. A few rules help:

• Hold it one to one, and not only once a year; regular check-ins through the year work better than one yearly talk.
• Talk about specific results and behaviour, with examples, not about the person's character.
• Let the employee speak first, and listen.
• End with agreed actions and a date to check.

Where it breaks. A fair process still cannot make a rating perfect. Rating is a judgement, and a rating tied to pay can make people hide problems. Managers who dread the conversation soften or delay it. That is feedback avoidance, and it is one of the main challenges.

For exams. Expect "name the error" from a short case, and the benefits and challenges as a list.

Go deeper for practitioners

A decision. Head office tells you to use a forced distribution on your team of eight: a few top ratings, most average, a few bottom. All eight met their targets. Forcing a bottom group means rating someone low for the sake of the curve. That can hurt trust and push people to compete rather than help each other. A calibration meeting, which tests ratings against evidence, may serve better than a fixed curve.

Links. Job evaluation (this unit) rates the job, not the person, and appraisal rates the person in the job. Compensation management (this unit) covers the incentive plans that appraisal results often feed.

The rater is the problem: halo, horn, lenient, strict, middle, recent.