What is surge capacity?
Surge capacity is a contact centre’s ability to take on far more work than normal, quickly, when something unexpected happens. It covers two things: the capacity itself (extra agents, automated answering, telephony headroom) and the plan that says where that capacity comes from, who switches it on and how fast.
The term is related to, but different from, a call avalanche. The avalanche is the event: a sudden flood of calls after a storm or outage. Surge capacity is what you have prepared to meet it. An organisation without a surge plan finds out how much capacity it has on the night, usually too late.
Why normal staffing models can’t provide it
Contact centres are rostered to forecast using workforce management and Erlang C calculations, with a margin for shrinkage and normal variation. That works for Monday mornings and billing cycles. It doesn’t work for an event that brings ten or twenty times normal volume within an hour.
Staffing permanently for the worst night of the year would leave most of those agents idle on every other night, with agent occupancy and cost per contact moving in the wrong direction. Surge capacity has to come from somewhere that is not sitting idle waiting for the storm.
Sources of surge capacity
Most resilient plans combine several sources, because each switches on at a different speed:
- Deflection. Proactive outage notifications, website banners and IVR announcements reduce how much capacity you need in the first place.
- Automation. An AI voice assistant can hold many conversations at once, so it adds capacity instantly and doesn’t need a roster change.
- Shared agents. A partner using blended resourcing can move already-trained agents onto your lines within minutes.
- Internal call-outs. Off-shift staff, on-call rosters and cross-trained back-office teams. Useful, but slower and dependent on people being reachable.
- Telephony headroom. Enough concurrent call paths and cloud capacity that callers reach a queue rather than an engaged tone.
How to plan surge capacity
- Model the event, not the average. Use your largest past events to estimate peak hourly volume, how long the plateau lasts and how much is repeat calling.
- Split the demand. Work out what share of calls is informational (status and restoration times) versus calls that need a person, such as hazard calls and vulnerable customers. Automation handles the first group; people handle the second.
- Define triggers. Agree in advance what queue length, wait time or event type activates each layer, and who can make that call at 2am.
- Prepare content. Scripts, knowledge articles, IVR messages and escalation rules should exist before the event, so surge agents give the same answers as your own team.
- Test it. Run exercises, check that surge agents can log in and see the right systems, and review after every real event.
- Agree the commercial model. Per-interaction pricing means you pay when surge capacity is used rather than for seats on standby.
Surge planning sits naturally alongside business continuity planning: the same storm that floods your phone lines may also affect your own staff, building or systems.