PMS

PMS stands for Performance Management System. It covers the whole cycle by which an organisation agrees what an employee is expected to achieve, tracks it, assesses it and acts on the result.

The abbreviation is used two ways and both are current. As a process, the PMS is the appraisal framework: goals, reviews, ratings, outcomes. As a product, a PMS is the software that administers that framework and holds the records.

Note the collision. Outside HR, PMS commonly means Property Management System in hotels and Project Management System in construction. In an HR context in India it is almost always performance management.

A complete cycle has five stages. Most organisations run three of them well and two badly.

The last weakness is the corrosive one. Where the increment budget is fixed and allocated first, the rating becomes a justification rather than an assessment, and employees work this out quickly.

Most Indian organisations use a four or five point scale. The choice of scale matters less than two decisions that sit around it.

The first is whether a distribution is enforced. Forced ranking, where a fixed proportion must fall in each band, produces comparability across managers and considerable resentment, particularly in small teams where the arithmetic guarantees someone is rated low regardless of performance. Guided distribution, where the shape is indicative rather than mandatory, is the more common compromise.

The second is calibration: managers reviewing proposed ratings together before they are released. This is the step that addresses the actual problem, which is that two managers assessing identical work will not produce the same rating. Calibration is unpopular because it takes a long meeting, and it is the single practice that most improves the credibility of the outcome.

An even-numbered scale removes the safe middle option and forces a direction. A five-point scale gives a genuine middle and collects most of the population there. Neither is wrong; the choice should follow whether the organisation actually wants to differentiate.

The recurring failures have little to do with the tooling.

• Goals set in January that nobody opens until December. The review then assesses the last six weeks, because that is what anyone remembers.
• Ratings with no consequence. If the increment is broadly uniform regardless of rating, the process has told everyone the rating does not matter.
• A form designed for HR rather than for the conversation. Twelve competency ratings on a five-point scale produce compliance and no insight.
• Managers who have never been trained to give critical feedback, delivering it once a year in a documented meeting that decides pay.
• Changing the framework every two years, so no employee ever sees the same system twice and no trend data survives.

The organisations that get value from a performance system tend to have done something unglamorous: shortened the form, increased the frequency of the conversation, and made the rating actually decide something.

The performance record is used well beyond the appraisal, and the connections are worth being deliberate about.

Compensation is the obvious one. Increment matrices, bonus multipliers and promotion eligibility all commonly key off the rating, which is why the rating has to be defensible at an individual level rather than only in aggregate.

The less obvious one is exit. Where an employee is being managed out for performance, the documented history is what makes the process fair and evidenced. A file showing three years of satisfactory ratings followed by an abrupt termination for non-performance is difficult to defend, and it is a common pattern precisely because managers avoid recording an honest low rating at the time.

The same record supports internal mobility, succession planning and identification of high potentials, all of which depend on ratings being comparable across managers. That returns to calibration, which is why it is worth the meeting.

Half-yearly suits most mid-sized organisations. The annual cycle is the default in India and it is under pressure for a reason: twelve months is long enough that the review assesses memory rather than performance.

The last row is where most organisations say they are and few of them are. Continuous feedback works where managers have been trained and the conversation is scheduled; where it is aspirational, it becomes no feedback at all, and the annual review returns with nothing recorded in between.

A practical middle position is to keep the annual rating for pay decisions, add one formal mid-year checkpoint that cannot be skipped, and keep the goal document open for revision at that checkpoint rather than treating it as fixed for the year.

What gets written in the review record matters more than the rating, because the record is what survives the manager.

A useful entry states what was expected, what happened, and what changes next period. A weak one records an adjective. The difference shows up when the employee moves teams, when a promotion case is made, or when performance becomes a formal issue and the file is the only evidence of what was said at the time.

• Write the assessment against the goals that were set, not against a general impression formed later.
• Record disagreement where it exists. An employee who contests a rating should be able to add their position to the record.
• Note what support was offered, particularly where performance was below expectation. A file showing a problem raised and nothing done about it is a weak position.
• Keep it retrievable. A review held in a manager's inbox is lost the moment they leave.

Managers avoid recording honest low ratings because it makes the following twelve months uncomfortable. The consequence arrives later, when someone with three years of satisfactory ratings is suddenly being managed out and nothing in the file supports it.