Workforce management (WFM) is forecasting contact volume, translating it into staffing needs, building schedules or rosters, and managing intraday adherence. It is the planning layer that decides how many agents, with which skills and languages, need to be working each hour.
At the desk it works against the cruellest demand curve in support. Travel volume is not driven by tidy day-of-week patterns but by external shocks weather, strikes, ATC outages, peak-season surges and cover has to span time zones and languages because travelers reach out around the clock. Done on spreadsheets it is slow and wrong; done well it forecasts volume closely and lifts schedule adherence sharply, and it routinely reveals that a desk has been carrying more headcount than demand actually needs. That is what keeps a lean travel help desk holding SLAs through chaos, and it is as true for an airline customer support desk, where a single disruption can flood the queue with time-critical tickets at once.
Edge cases: The most volatile input now is the AI resolution share a bot outage reroutes its entire volume back to humans, so the forecast has to model that failure mode. Overstaffing by even one full-time agent carries a real annual cost, while understaffing a disruption breaches SLAs exactly when it matters most, so the rollout order that works is forecasting first, scheduling second, adherence third, intraday last.
Common questions
How do we forecast for disruptions we can’t predict? You staff for the shape, not the event build flex capacity and on-call tiers so the desk can absorb a spike, rather than trying to predict which storm hits when.
How do we staff overnight for APAC suppliers? Schedule the supplier-chase work to the supplier’s local business hours and the traveler-facing cover to traveler demand they are two different curves and staffing them as one leaves gaps