Variable Pay

Why does the plan document matter more than the formula?

Because the formula is rarely what is disputed. Arguments about variable pay concentrate on whether a target was met, whether the measurement was the one intended, whether somebody who joined or left partway through qualifies, and whether a discretionary reduction was applied fairly. A formula answers none of those and cannot be made to. A plan document stating the measure, its source, the period, the eligibility conditions and who decides answers all of them in advance, which is the only moment they can be answered without somebody feeling they lost an argument to a person who benefited from winning it.

What must a plan settle before the period starts?

What is being measured and where the figure will come from; over what period performance is assessed and when payment follows; who qualifies and what happens to people whose employment changes partway through; whether any element is discretionary and who exercises it; and what happens if the measure turns out to be unavailable or misleading. Settling these afterwards is possible but expensive, because by then everybody involved holds a view about the outcome and no view on the table is disinterested, including the view of the person writing the clarification.

Why is discretion harder to operate than it looks?

Discretion is included so a plan can respond to circumstances a formula cannot anticipate, which is a reasonable aim and frequently the right design. What it produces in practice is a payout the recipient can neither predict nor check, and an obligation on whoever exercises it to defend a decision that has no stated basis to fall back on. If discretion is retained, name who holds it, describe the factors it will weigh, and expect to give reasons every time. A discretionary element nobody can explain is read as arbitrary, and over enough cycles it becomes so.

What is actually being paid for?

The measure has to be something the recipient influences and something the organisation can produce reliably, and those two conditions conflict more often than plan designers expect. Company-level results are reliable and barely influenced by any individual below the top of the organisation; individual measures are influenceable and are frequently assembled by hand from sources never intended to settle a payment. A measure drawn from a system that also serves as the record of the work is stronger than one reconstructed at period end, because nobody is reconstructing it with the payout already in view and the arithmetic already half-known.

Composite plans with several weighted measures are common and worth limiting deliberately. Past a small number of components the recipient can no longer predict the effect of their own behaviour, which removes the incentive the plan was built to create, while the administration grows faster than the benefit does. Where individual measures come from a [performance management system](/performance-management-software), the same rating that drives development conversations begins carrying a payment, and that changes how candidly those conversations are held by both parties - a real trade-off to be accepted with open eyes rather than an argument against doing it.

When is it earned and when is it paid?

Earning and payment are separate events, and the gap between them is where most eligibility disputes originate. A payment relating to one period, calculated after it closes and paid some periods later, raises an immediate question about anyone who leaves in between: did they earn it during the period, or does entitlement depend on being employed on the payment date? Both answers are in use across the market and both are defensible on their own terms. Only one can be used consistently inside a single organisation, and it has to be written down before anybody resigns and asks.

The interval also affects how the payment is experienced when it finally arrives. A large amount landing in a single run has a different effect on that period's deductions than the same amount spread across the year, and employees notice immediately and ask why. Flagging the payment period in advance, and showing the amount as its own line in the run, prevents the statement from being the first place somebody learns their incentive was paid and how much of it they will keep. Where the amount is substantial, saying so beforehand is more considerate than explaining afterwards to somebody already disappointed.

How are joiners, leavers and movers treated?

Three populations break any plan written with only continuous employees in mind. Somebody joining partway through has performed for part of the period; a plan ignoring this pays nothing and quietly discourages mid-year hiring, while one paying in full pays for work that did not happen. Pro-rating from the joining date is the common answer and needs to be stated rather than improvised at the moment it first arises, because the first improvisation becomes the precedent everybody cites afterwards. Somebody leaving raises the earning-versus-payment question, with more at stake because the relationship is already ending.

Movers are the population plans most often forget entirely. Anyone changing role, team or manager during the period may have performed against two different sets of measures, and whoever inherits them at period end usually has no visibility of the earlier half and no incentive to go looking for it. The workable rule assesses each portion against the measures in force at the time and combines the results, which requires that the earlier assessment was recorded while it was still current and while the previous manager still cared. Deciding this afterwards produces an outcome resembling whatever suits the person deciding.

How does variable pay quietly become fixed?

A payout arriving every period at a similar level stops being read as contingent, whatever the plan says. Employees plan around it, count it when comparing offers elsewhere, and treat a reduction as a pay cut rather than as the plan operating exactly as designed. That drift is caused less by generosity than by targets set once and never revisited, or by a discretionary element consistently exercised upward to avoid a difficult conversation this quarter. The plan document still says the payment is conditional; nobody in the organisation believes it any more, including the people administering it and the managers defending it.

Two habits slow the drift without requiring a redesign. Revisit the measures each cycle so they reflect current priorities rather than the ones in force when the plan was drafted, and be willing to let a payout differ when performance differs - which is markedly easier the first time than the third, because each unchallenged cycle raises the expectation further. It also helps to keep the fixed and variable elements visibly separate in the [payroll software](/payroll-software) and on the statement, so the conditional part is presented as conditional every period instead of merging into something that reads as salary.