Back-office & blended work

Free visual lesson · about 5 minutes · short quiz at the end

ccPlanning academy · advanced

Back-office & blended work

Not everything is a call waiting to be answered in 20 seconds.

The big idea

Deferrable work is a different planning problem.

Inbound calls are real-time: answer now or lose them. Back-office work — case processing, complaints, applications, emails, fulfilment — can wait hours or days. That single difference changes the maths, the metrics and the staffing approach entirely. Applying call-centre Erlang to it is a category error.

Different maths

Throughput, not Erlang.

You don’t need enough people for the peak instant — you need enough to clear the work within its deadline. Planning is about capacity over a window: items in, items processed, backlog managed against an SLA. It’s closer to a factory production line than a phone queue.

The backlog is your friend

Deferrability is a smoothing tool.

Because the work can wait, you can hold a controlled backlog through a peak and clear it in the trough — staffing to the average, not the peak. That’s far more efficient than live channels, which can’t bank demand. The skill is keeping the backlog inside its SLA, not at zero.

Different metrics

Age and SLA, not service level in seconds.

The right measures are the age of the oldest item, the percentage cleared within the promised time (e.g. 95% in 48 hours), and backlog size vs capacity. “80% in 20 seconds” is meaningless here — forcing live-channel metrics onto deferrable work drives the wrong behaviour.

The blending trap

“Do email between calls” rarely works.

Blending live and deferred work on the same agents sounds efficient, but the live channel always wins — agents fill every quiet second with calls, and the back-office work silently ages into an SLA breach. Blended models need protected, ring-fenced time for the deferred work, or it never gets done.

Doing it well

Plan the two streams together, staff them honestly.

Forecast back-office arrivals and processing capacity separately, set realistic SLAs, and decide deliberately how to blend — dedicated case teams, ring-fenced hours, or using deferred work as genuine fill for live troughs. Just don’t pretend it’s free capacity hiding between calls.

Why blending quietly fails

“Do the cases between calls”

It sounds like free capacity: agents handle email and case work in the gaps between calls. But the phone always wins — every quiet second fills with a ringing line, and the back-office pile silently ages past its 48-hour SLA while the dashboard looks fine.

The deferred work isn’t getting done in the gaps; it’s being starved by them. Blended models need ring-fenced time for cases, or they breach.

The takeaway

Deferrable work: throughput, SLAs, protected time.

Back-office work waits, so plan it like a production line — capacity to clear the work within its deadline, staffed to the average with a managed backlog, measured by age and SLA. Blend with live channels deliberately and protect the deferred time, or it ages into breach.

That’s the track — now test yourself ↓

1 / 8

Slides done? Here’s the same idea in a bit more depth — the part worth keeping.

In depth: not everything is a call to answer in 20 seconds

Inbound calls are real-time — answer now or lose them — but back-office work like case processing, complaints, applications, emails and fulfilment can wait hours or days. That single difference changes the maths, the metrics and the staffing approach entirely, and applying call-centre Erlang to deferrable work is a category error that wastes money or breaches SLAs.

Throughput, and the backlog as a friend

You don’t need enough people for the peak instant; you need enough to clear the work within its deadline. Planning is capacity over a window — items in, items processed, backlog managed against an SLA — closer to a factory production line than a phone queue. And because the work can wait, deferrability is a smoothing tool: you can hold a controlled backlog through a peak and clear it in the trough, staffing to the average rather than the peak, which is far more efficient than live channels that can’t bank demand. The skill is keeping the backlog inside its SLA, not at zero. The metrics follow from this — the age of the oldest item, the percentage cleared within the promised time, and backlog versus capacity — because “80% in 20 seconds” is meaningless and forcing it onto deferrable work drives the wrong behaviour.

The blending trap

“Do email between calls” sounds efficient and rarely works: the live channel always wins, agents fill every quiet second with calls, and the back-office work silently ages into breach. Blended models need protected, ring-fenced time for the deferred work, or it never gets done. Done well, you forecast back-office arrivals and processing capacity separately, set realistic SLAs, and decide deliberately how to blend — dedicated case teams, ring-fenced hours, or deferred work as genuine fill for live troughs — just without pretending it’s free capacity hiding between calls.

The principle to remember: deferrable work means throughput, SLAs and protected time. Plan it like a production line — capacity to clear the work within its deadline, staffed to the average with a managed backlog, measured by age and SLA — and blend with live channels deliberately, protecting the deferred time or it ages into breach.

Back-office & blended work