Reliability index

What it measures

Reliability measures the quality of the behaviour observed, separated from how long it has been measured. It answers a specific question: can the manager trust that this employee will be where they should be, when they should be and for the time planned?

That separation is deliberate and important: someone who has been with the company for a month does not carry a low reliability simply because they have little track record. What is adjusted in that case is the pattern's confidence level, which indicates how solid the calculation is.

A reliable employee is not necessarily the most brilliant or the most productive on complex tasks. They are the one whose behaviour is predictable: they arrive on time, complete their workday and generate no unexpected absences. That consistency has high operational value, especially where shift planning depends on each person being present.

Reliability is also one of the three main indices of the Operations dashboard, alongside site effect and dependency, and it is read at organisation, site and person level. See The Operations dashboard.

Which behaviours influence the index

The reliability index is built from the accumulated analysis of several behaviours:

Punctuality on arrival. The system compares the real entry time with the one planned in the effective calendar. Repeated late arrivals reduce the index in proportion to their frequency and to the size of the deviation.

Completing the full workday. Arriving on time is not enough; the system also evaluates whether the employee completes the planned working time. Frequent early exits affect the index in a similar way to late arrivals.

Unjustified absences. An absence with a medical certificate or approved by the manager is treated differently from an absence that is not formally recorded on the platform. Accumulated unjustified absences are one of the factors with the greatest negative weight in the index.

Schedule deviations. The system looks not only at whether the employee arrives late or leaves early, but at the size of those deviations over time. Small, sporadic deviations have less impact than large, recurring ones.

How it is presented

Reliability is not published as a mark out of 100 or as a ranking. It is shown with its variation against the previous period —which usually says more than the absolute value— and with its maturity state:

When volume is lacking, the dashboard indicates how many records are needed and how many there are. Reliability that does not appear does not mean low reliability: it means nothing can be asserted yet.

How an employee improves their index

The reliability index is a weighted average that evolves over time. Because it is based on accumulated behaviour, it does not change instantly, but it does respond progressively to sustained improvement.

An employee whose index has fallen can recover it by maintaining consistent behaviour over weeks or months: arriving within the punctuality margin, completing their workdays and avoiding uncommunicated absences. The system weights more recent data more heavily, which means that a real, sustained improvement is reflected in the index within a reasonable time.

How the manager uses it

For team leads, the reliability index has several practical uses:

Planning critical shifts. When a particularly important shift has to be covered, it is useful to assign it to employees with a high index, whose presence is more predictable.

Early identification of problems. A sustained fall in the index of an employee who used to be consistent may indicate a personal, motivational or role-fit problem worth addressing before it worsens.

Team comparison. A team's or branch's average index makes it possible to compare collective reliability between different operational units.

Grounding conversations. When the manager needs to talk to an employee about their absences or punctuality, the index history offers an objective data basis that makes the conversation easier and fairer for both parties.

Limitations of the index

The reliability index measures behaviour, not causes. A low score may be due to work-life balance problems, a badly distributed workload, conflicts within the team or personal circumstances the system does not know about. The index identifies the pattern; the responsibility for understanding the reason and acting accordingly lies with the manager.

Nor does it measure the quality of the work or performance on tasks. An employee with high reliability may perform poorly in their responsibilities. For a fuller picture of the employee, also consult the operational efficiency index and their operational assessment.