Failure demand: the contacts you should never have received
Two kinds of demand, and only one you want
Not all contact volume is created equal. Some of it is the work the centre exists to do — a customer wants to buy, to change something, to ask a question they could only ask a person. That is value demand. The rest is something else: contacts that exist only because the organisation failed the customer somewhere upstream. A bill that was wrong. A promise that wasn’t kept. A website that wouldn’t let them do it themselves. A first call that didn’t resolve, so they called again. That is failure demand — a term coined by the systems thinker John Seddon — and in a lot of operations it is a startlingly large share of the total.
For the planner this distinction matters enormously, because the two behave completely differently. Value demand is roughly a function of how many customers you have and what they need. Failure demand is a function of how well the rest of the business works — and unlike value demand, it can be removed. Treating the two as one undifferentiated volume is how a centre ends up efficiently, accurately, expensively staffing work that should not exist.
Why planners quietly legitimise it
Here is the uncomfortable part. Failure demand arrives in the same queues as value demand, gets handled by the same agents, and shows up in the same volume history. So the planner does what planners do: forecasts it, staffs for it, and reports the service level on it. Nothing in the standard process distinguishes a “why is my bill wrong again” call from a “I’d like to upgrade” one. And by forecasting failure demand faithfully and resourcing it well, the planning function quietly legitimises it — turning a symptom of organisational failure into a permanent, budgeted line of business that everyone has stopped questioning.
The efficiency drive makes it worse. Squeeze AHT, lift occupancy, automate the easy contacts, and you become superbly good at processing failure demand cheaply — which removes the very cost pressure that might have forced someone to fix the upstream cause. An operation can hit every target on a wall of work it should not be doing.
How to find it
You cannot manage what you do not name, so the first move is to make failure demand visible. That means contact-reason coding designed around a single question: did this contact exist because we failed to do something, or do it right, for this customer? Not “billing” versus “sales,” but “chasing something we promised,” “correcting something we got wrong,” “asking again because last time didn’t resolve.” A short period of listening to calls with that lens is usually enough to shock a leadership team — in service operations the share is frequently a third or more, sometimes much higher. The demand decomposition work is the structured version of this; failure demand is the cut of it that matters most.
Repeat contacts are the easiest slice to see, because first contact resolution already measures them: every avoidable repeat is failure demand the customer is generating because the first attempt didn’t land. If you track nothing else, track that.
The planner’s honest stance
The planner cannot remove failure demand — its causes sit in product, process, billing, IT and policy, not in the contact centre. But the planner is often the only person positioned to see it across the whole operation, to size it, and to put a number on what it costs. That is a powerful role to play. Forecast the volume honestly — you still have to staff tomorrow’s reality — but split the forecast so the failure component is visible, attach the cost, and take it to the people who own the upstream cause. A pound figure on “avoidable contacts” moves conversations that an abstract appeal to quality never will; the cost-per-contact calculator turns the volume into that figure.
The mindset shift is the whole point. Most planning energy goes into meeting demand more efficiently. The highest-leverage move available to a contact centre is usually not to handle the work better but to stop receiving it — and the planner who can show, in money, how much of the queue should never have existed is the one who gets that conversation started.
Pair this with decomposing demand by call reason, first contact resolution, and why deflection raises AHT.