On many contact centre floors, Monday morning has a familiar shape: the roster says 200 agents, around 176 log in, and the real-time analyst spends the first hour moving breaks and cancelling coaching. The same thing happens, more sharply, on the first working day after Diwali, Pongal or Eid. It is tempting to treat this as a discipline problem and respond with warnings. Sometimes that is right. More often, the pattern is telling you something about planning, policy or the job itself, and the response should start with reading the data properly.
Planned and unplanned absence are different problems
Workforce planners split absence into two kinds. Planned absence, such as approved leave and scheduled training, is known in advance and can be built into the roster. Unplanned absence, including sick days, unapproved leave and no-shows, is known only on the day. Both are part of shrinkage, but only unplanned absence damages service level directly, because the roster was built assuming those people would be present.
A common way to measure it is unplanned absence hours divided by scheduled hours, over a week or month. Definitions vary between organisations: some count half-day absences, late logins beyond a threshold, or leave applied on the same morning as unplanned. Whatever the definition, it must be consistent before patterns mean anything.
Patterns worth looking for
Slicing absence data by a few dimensions usually reveals more than the headline rate:
- Day of week: if Monday runs at 12% unplanned absence and midweek at 6%, that is a planning input, not just a behavioural issue.
- Adjacent to holidays and weekly offs: absences touching a weekend or festival suggest people are extending breaks, often because they need to travel to home towns.
- Shift: night and early-morning shifts often show higher absence, linked to transport, sleep and family obligations.
- Tenure: very new agents often have high absence that precedes resignation; it can be an early attrition signal.
- Team: if one team leader’s team has twice the absence of others on the same process and shift, look at that team’s management and climate.
The Bradford factor, and its limits
The Bradford factor is a formula, widely used in UK workplaces and sometimes in Indian BPOs, that scores frequent short absences as more disruptive than occasional long ones. It is calculated as S × S × D, where S is the number of separate spells of absence and D is the total days absent in a period, usually a rolling year.
Consider three agents, each absent for ten days in a year:
- One spell of 10 days, perhaps an illness: 1 × 1 × 10 = 10.
- Five spells totalling 10 days: 5 × 5 × 10 = 250.
- Ten single-day absences: 10 × 10 × 10 = 1,000.
The logic is operational: ten unannounced single days are harder to plan around than one known block. The factor is useful for spotting patterns that a plain day count hides.
Its weakness is that it is blind to reasons. An agent managing a chronic condition, a disability or caring responsibilities may have frequent short absences for entirely legitimate reasons, and a mechanical trigger based on the score can be unfair and, depending on the circumstances and applicable law, discriminatory. Used responsibly, a high Bradford score is a prompt for a supportive conversation, not an automatic disciplinary step.
Planning around predictable spikes
If Monday absence is reliably 12%, scheduling as if it were the weekly average is choosing to be short. To have 170 agents present on a Monday at 12% absence, you need to roster about 170 ÷ 0.88, or roughly 194 agents. At a midweek rate of 6%, about 181 would do. Building day-specific absence assumptions into the roster turns a recurring crisis into a planned cost.
For festivals, the patterns are even more predictable. Practical approaches include:
- Opening leave windows early so agents can plan travel legitimately, with fair rotation across years for the most popular dates.
- Mapping the workforce by home state, since a site in Bengaluru or Chennai may have large groups who celebrate different festivals at different times.
- Agreeing with the client, in advance, on volume expectations and staffing for festival weeks.
- Planning the return day: many agents travel back overnight, so the first day after a long festival break is often worse than the festival itself.
Finding the cause before choosing the fix
Absence that clusters around weekends can reflect a rigid weekly-off policy. If agents never get a Saturday or Sunday off, they will take one. Rotating weekly offs or allowing shift swaps often reduces Monday absence more than any warning letter.
High absence in one shift may point to transport problems. Absence concentrated in one team often reflects the team leader relationship. Rising absence across a process can be a sign of burnout from sustained high occupancy and overtime.
Attendance incentives, such as a bonus for full monthly attendance, can help, but they have a cost: they encourage people to come in sick, and they penalise the one legitimate absence as heavily as a pattern of avoidable ones. Many floors get better results combining modest recognition for attendance with flexibility, such as easier shift swaps and short-notice leave within a quota, so that people who need a day can take it in a way the roster can absorb.
The underlying principle: treat predictable absence as a planning variable, unpredictable individual absence as a conversation, and team-wide spikes as a signal about the work environment.
Written by the RP Strativa Team. Rates and headcounts are illustrative.
Leave a Reply