The schedule that’s right for the steady state and wrong for the peak
Excellent 35 weeks, broken 17
The schedule was built against the operation’s steady-state demand pattern. For about 35 weeks of the year, the pattern matches and the schedule delivers. For the other 17 — the December peak, the marketing-campaign weeks, the post-holiday rebound, the regulatory deadline, the summer trough — the steady-state schedule is wrong. The operation responds with emergency overtime, last-minute outsource bursts, and a real-time team that spends those 17 weeks firefighting.
The failure isn’t that the steady-state schedule is wrong. It’s that the operation hasn’t designed for the predictable variation. The seasonal events that catch operations every year are, by definition, predictable. The honest move is to design the steady-state schedule with peak absorption built in, rather than rebuilding the schedule every quarter.
The four design moves that absorb seasonal drift
1. A deliberate flex layer. A small group of agents (often 5–10% of headcount) whose contractual hours are above the baseline use, with the variable hours scheduled into peak weeks. The layer absorbs 50–70% of typical peak variation without overtime. The contract conversation up-front is harder than emergency overtime; the cost-per-hour is significantly lower.
2. Agreed overtime rules. Overtime that’s pre-agreed at a defined rate and a defined cap for named peak weeks is operationally clean. Overtime that’s a panic response to an unforeseen peak is expensive and demoralising. Operations that pre-commit overtime budgets to known peaks find both planners and agents prefer it.
3. A small annualised-hours pool. See annualised hours. A subset of the workforce on annualised contracts can flex significantly across the year — light hours in the summer trough, heavy hours in December. The administration is real but the absorption capacity is substantial.
4. The standing outsource overflow. A pre-agreed outsourcer relationship that can pick up named peak weeks. The unit cost is higher than in-house but the SL absorption is reliable and the contractual relationship is in place before the peak hits. Operations that arrange this every year rather than every peak find the outsourcer pricing improves over time.
The events worth pre-mapping
Most operations are caught by the same calendar events every year. A single annual exercise — the seasonal-events map — catches most of them.
Christmas and pre-Christmas. Volume spikes from mid-November through to the new year. The dates are known. The pattern is consistent.
Black Friday / Cyber Monday. Late November. Sector-specific in intensity but reliable in timing.
Easter and the half-term weeks. Education and family-services operations have a different curve; retail tends to peak around the Easter weekend.
Bank holidays. The pre- and post-bank-holiday pattern is reliable; see bank-holiday volume patterns.
Regulatory deadlines. Tax year-end, FCA Consumer Duty attestation windows, switching windows, year-end accounts. The dates are public.
Marketing peaks. Annual campaigns, product launches, end-of-promotion pushes. The marketing team knows the calendar months in advance.
The World Cup or similar four-yearly events. Forecasted years ahead. See planning during the World Cup.
Briefing leadership on the design choice
The conversation with operations leadership is honest: the schedule is designed for steady state with named peak exceptions. We can either invest in the flex layer and the agreed overtime rules now, or we can spend two months a year firefighting. The trade-off is between predictable cost up-front and unpredictable cost (and SL impact) in the peak. Most operations leadership, when the trade-off is named explicitly, choose the predictable cost. Most planning teams don’t name the trade-off explicitly, so the choice gets made by default in favour of firefighting.
The diagnostic for whether your operation is doing this well
Three questions. Do your peak weeks have a named operational plan in place six weeks in advance? If they don’t, you’re scheduled for steady state and you’ll firefight the peaks. Is your overtime spending heavily concentrated in 4–8 weeks of the year? If yes, your schedule isn’t absorbing seasonal variation. Does your peak-week SL miss substantially exceed your steady-state SL miss? If yes, the schedule design is the problem, not the forecast.
Conclusion
Seasonal drift is one of the most predictable scheduling failures and one of the most under-managed. The fix is structural: design the steady-state schedule with peak absorption built in, pre-map the seasonal events, and name the design choice to operations leadership before the peak rather than after. Operations that do this handle their peak weeks calmly; operations that don’t spend two months a year explaining why service collapsed in a week the calendar told them about months in advance.
Next in the series: The schedule that ages out.
Pair this with peak season planning, the annual planning calendar, and annualised hours.