Blending and Deferred Workload
Blending and Deferred Workload is the practice — and the staffing math — of having the same agents handle time-constrained inbound work (calls, chat) and non-time-constrained deferrable work (email, back-office tasks, outbound) so that the idle capacity left over from staffing inbound to a service level is used to clear deferred work rather than wasted. It is one of the highest-leverage efficiency levers in workforce management, and it is the specific sense of "blending" in which inbound is combined with channels that have no second-by-second deadline. This page covers the capacity logic; for handling multiple real-time channels and chat concurrency, see Multi-Channel and Blended Operations, and for the deadline side of deferred work, see Back Office and Knowledge Worker Workforce Management.
Time-constrained versus non-time-constrained work
The two kinds of work have fundamentally different deadlines. Inbound calls and chat must be answered within seconds, so they are staffed to a service-level target and cannot wait. Email, back-office processing, and outbound campaigns have deadlines measured in hours or days; they are deferrable. This difference is what makes blending possible: deferrable work can be parked and picked up whenever inbound demand dips, acting as a flexible buffer that absorbs spare capacity.
The idle-time fill model
Any inbound queue staffed to a service level necessarily runs at occupancy below 100% — by Erlang, some idle time is required to absorb the randomness of arrivals.[1] That idle time is not waste to be eliminated; it is the reserve that protects service level. But it is also an opportunity: in the troughs of the day, and in any period that is over-staffed relative to inbound demand, agents have time that deferrable work can fill without adding headcount. Blending converts that otherwise-idle reserve into completed deferred work, raising effective utilization while inbound service level is held.[2]
The key discipline is that inbound takes priority: when a call arrives, the agent leaves the deferred task and takes it (preemption), so inbound service is protected and deferred work flexes around it. The diagram shows the pattern — deferred work fills the gap between inbound demand and scheduled capacity, while a sliver of reserve is kept genuinely idle to handle arrival randomness.
Why you cannot blend away all idle time
The common error is to treat all sub-100% occupancy as slack to be filled. It is not. The idle fraction in an inbound plan is the safety margin against random arrivals; consume it entirely with deferred work and inbound service level collapses the moment a burst arrives. The blendable capacity is therefore only the portion of idle time beyond the reserve the service-level target requires — larger in low-volume periods (where the square-root safety cushion is a big fraction of staffing) and smaller at peaks. This is why the pooling and blending benefit is greatest overnight, on small queues' quiet stretches, and in large agent pools where randomness averages out.
Staffing implications
- Deferred work may need little dedicated headcount. If inbound idle time is sufficient and deadlines are loose, deferrable volume can be cleared largely within the blend, rather than sized as a separate team — provided the plan does not double-count the shared agents.
- Or the same headcount clears more total work. Equivalently, blending lets a given inbound roster absorb a deferred backlog, improving cost per unit of total work.
- Plan deferred work as a buffer, not a silo. The capacity plan should model deferrable work as flexible fill against inbound idle time, with a minimum guaranteed allocation so it does not starve, rather than as an independent staffing requirement.
Cautions
- Deferred work can starve. On a queue that is busy all day, there is no idle time to blend into; without a protected minimum allocation, the deferred backlog grows unbounded. Deadlines on deferred work must still be managed (see Back Office and Knowledge Worker Workforce Management).
- Switching costs and quality. Frequent interruption between a complex back-office task and inbound calls carries a real context-switching cost and can degrade quality on both; the blend ratio is a managed real-time decision, not a free lunch.[3]
- It does not rescue an understaffed inbound plan. Blending uses surplus capacity; if inbound is under-staffed there is no surplus, and pushing deferred work in only worsens service. Blending is an efficiency lever, not a staffing substitute.
Maturity Model Position
In the WFM Labs Maturity Model™, how an operation handles deferrable work is a clear efficiency-maturity signal.
- Level 1–2 (Emerging / Foundational) — inbound and deferred work are staffed as separate silos; idle inbound time is wasted while a separate team works email, and the double-counted plan is over-staffed in total.
- Level 3 (Progressive) — deferrable work is blended into inbound idle time with inbound priority and a protected minimum, raising utilization without harming service level.
- Level 4–5 (Advanced / Pioneering) — blend ratios are managed dynamically against real-time inbound conditions and deferred-work deadlines, and the capacity plan models deferrable work as a flexible buffer across the pool.
See also
- Multi-Channel and Blended Operations
- Back Office and Knowledge Worker Workforce Management
- Occupancy
- The Occupancy Trap
- Multi-Skill Pooling and the Double-Counting Trap
- Erlang X
- Square Root Staffing Law
References
- ↑ Gans, N., Koole, G., & Mandelbaum, A. (2003). "Telephone Call Centers: Tutorial, Review, and Research Prospects". Manufacturing & Service Operations Management, 5(2), 79–141.
- ↑ Bhulai, S., & Koole, G. (2003). "A Queueing Model for Call Blending in Call Centers". IEEE Transactions on Automatic Control, 48(8), 1434–1438. doi:10.1109/TAC.2003.815038.
- ↑ Deslauriers, A., L'Ecuyer, P., Pichitlamken, J., Ingolfsson, A., & Avramidis, A. N. (2007). "Markov Chain Models of a Telephone Call Center with Call Blending". Computers & Operations Research, 34(6), 1616–1645.
