Quarter-End Demand Spike at PayrollBridge
PayrollBridge provides a cloud payroll platform to 1,100 small businesses. Demand is usually about 900 support contacts per week, but in the three days before each quarter-end it rises to 3,200. Analysis of the last two quarter-ends shows 65% of contacts are password or MFA issues, 22% are payroll-run errors, and 13% are reports or other requests. The support team has 14 agents, a four-hour SLA for payroll-run errors and a next-business-day target for password issues. No extra headcount is approved for the next quarter-end, which is eight weeks away. A self-service password reset exists but only 18% of users have enrolled. The CFO has warned that any payroll error affecting pay is unacceptable; the COO wants to stop the 'ticket tsunami' without harming clients. You are asked to prepare a demand management plan.
1. How would you segment and classify the demand before deciding whether to add capacity, deflect contacts or reshape demand?
Read the worked answer
Segment before acting. The quarter-end spike is predictable, but the 3,200 contacts are not one demand type. The 65% password/MFA group is largely avoidable failure demand: low value, repeatable and suitable for self-service and pre-quarter-end enrolment. The 22% payroll-run errors are value demand with high financial risk and need skilled, prioritized capacity. The 13% reports can be deflected or scheduled. Use historical arrival patterns and service times to model queue pressure; do not average across segments. The decision is therefore not 'add 14 agents or not' but 'shape the low-value demand, protect capacity for payroll errors, and schedule the rest'. Without segmentation, a headcount request would be sized on the wrong workload and the root cause would remain.
2. Design a plan for the next quarter-end with no new headcount and an eight-week lead time.
Read the worked answer
Eight-week plan, no new headcount: weeks 1-2 run an MFA/password-reset enrolment campaign with a target above 70% of active clients and pre-quarter-end validation checks; weeks 3-4 publish payroll checklists and add a triage script that routes payroll-run errors to a dedicated skilled queue; weeks 5-8 staff shifts to match arrival patterns, add callbacks for password issues and knowledge articles for reports. Protect the four-hour payroll-error SLA and keep a daily war room in the three spike days. Measure enrolment, self-service success, queue wait by segment and repeat contacts. If payroll-error volume still exceeds capacity, escalate with evidence rather than silently letting the SLA fail.
3. How would you verify that demand shaping improved performance rather than simply suppressing contacts?
Read the worked answer
Check both leading and lagging signals. Leading: enrolment, self-service completion, deflection of password contacts, repeat-contact rate. Lagging: payroll accuracy, four-hour SLA attainment, abandonment, CSAT and client complaints. Compare a cohort of clients exposed to the campaign with a matched group, and sample deflected contacts to confirm they were genuinely resolved, not suppressed. A drop in contacts with a rise in repeat calls or payroll errors is not success. Also track agent workload and overtime to see whether demand was shifted rather than hidden. Report the trade-offs explicitly to the COO and CFO.