WFM guideIntraday

Contact centre intraday escalation playbook

Q: In what order should you deploy intraday levers when service level is dropping?

Deploy levers in order of speed, cost, and customer impact — fastest, cheapest, and least customer-visible first: (1) Reallocate breaks and offline activities — push discretionary breaks, meetings, training, and coaching out of the affected window. This is free, instant, and invisible to customers. (2) Stop discretionary off-phone work — pause outbound callbacks, admin time, and project work; return those agents to the live queue. (3) Skill-flex — route multi-skilled agents from a quieter queue to the queue under pressure. (4) Call in overtime — offer immediate overtime to off-shift agents (slower to take effect, and it costs money). (5) Overflow — route contacts to another site, team, or outsourcer if that capability exists. (6) Customer communication — as a last resort, set expectations with customers (queue messaging, estimated wait, callback offer) to manage the experience of the wait you cannot eliminate. The principle: exhaust the free, instant, invisible levers before reaching for the slow, expensive, or customer-visible ones. A team that calls in overtime before reallocating breaks is spending money to solve a problem a free lever would have fixed.

When the queue is building, the worst response is an ad-hoc scramble. A pre-agreed playbook turns panic into a sequence: defined triggers, defined tiers, and a defined order for the levers — exhausting the free, instant, invisible ones before the expensive or customer-visible ones.

The lever order: cheap, fast, invisible first

Reallocate breaks & offline activityFree · instant · invisible to customers

Stop discretionary off-phone workFree · fast · invisible

Skill-flex agents from quieter queuesFree · fast · invisible (if multi-skilled cover exists)

Call in overtimeCosts money · slower to take effect · invisible

Overflow to another site / outsourcerCosts money · needs prior capability · semi-visible

Customer communication (manage the wait)Last resort · customer-visible · manages rather than fixes

A team that calls in overtime before reallocating breaks is spending money to solve a problem a free lever would have fixed. Always work down the list, not in from the middle.

The four escalation tiers

Tier 0 — Monitor

Real-Time Coordinator (RTC)

Trigger

SL within tolerance; oldest contact within target; no queue build. Normal operating state.

Levers to deploy

No action beyond standard real-time monitoring. Watch the leading indicators (queue depth, oldest contact, agents available) for early signs of deterioration.

Tier 1 — Soft response

Real-Time Coordinator (RTC)

Trigger

SL dipping below target for 1–2 intervals, OR oldest contact approaching 2× the SL target wait, OR a modest queue build. Early deterioration, recoverable with free levers.

Levers to deploy

Reallocate breaks out of the affected window. Pause discretionary off-phone activity (admin, project work). Defer non-urgent coaching/1:1s. All free, instant, invisible to customers.

Tier 2 — Active intervention

RTC + Operations Manager (informed)

Trigger

SL below target for 3+ intervals despite Tier 1, OR a sustained queue build, OR oldest contact beyond 3× the SL target. Tier 1 levers are insufficient.

Levers to deploy

Skill-flex multi-skilled agents from quieter queues. Pause all outbound/callback work. Call in immediate overtime from off-shift agents. Brief team leaders to maximise floor presence and minimise unscheduled off-phone time.

Tier 3 — Major escalation

Operations Manager + Senior Leadership

Trigger

SL in serious breach with no recovery trajectory, OR a sudden volume spike beyond all available capacity, OR an event (system outage, major incident) causing a structural shortfall. Internal levers exhausted.

Levers to deploy

Activate overflow to another site / team / outsourcer if available. Deploy customer communication (queue messaging, wait-time expectations, callback offers). Escalate to senior operations leadership. Consider invoking the volume-spike or BCP plan if the trigger is an event.

Why a pre-agreed playbook beats real-time judgement

Speed

In a queue crisis, every interval of delay deepens the deficit (queues build non-linearly). A playbook removes the deliberation step — the RTC acts on a pre-agreed trigger rather than stopping to decide what to do and who to ask.

Authority

The playbook pre-authorises the RTC to deploy Tier 1–2 levers without seeking sign-off. Waiting for a manager's approval to reallocate breaks wastes the window in which the break reallocation would have helped.

Cost discipline

Defining the lever order in advance prevents the expensive reflex (call in overtime immediately) when a free lever (break reallocation, pausing admin) would have worked. The playbook enforces the cheap-first sequence.

Consistency

Different RTCs respond to the same situation the same way. Without a playbook, the response depends on who is on shift — some escalate too early (wasting cost), some too late (letting SL collapse).

Intraday escalation questions

In what order should you deploy intraday levers when service level is dropping?

Fastest, cheapest, least customer-visible first: (1) reallocate breaks and offline activity (free, instant, invisible); (2) stop discretionary off-phone work and return those agents to the queue; (3) skill-flex multi-skilled agents from quieter queues; (4) call in overtime (costs money, slower to take effect); (5) overflow to another site/team/outsourcer if that capability exists; (6) customer communication (queue messaging, wait expectations, callback offers) as a last resort to manage the wait you cannot eliminate. The principle is to exhaust the free, instant, invisible levers before reaching for the slow, expensive, or customer-visible ones — calling in overtime before reallocating breaks spends money on a problem a free lever would have fixed.