Map a real month, not a tidy one
List demand in the words your customers actually use, not your ticket categories. Translating demand into internal categories is how organisations lose the ability to see it. Then answer the three questions above for each line.
What you need, and what you will leave with
An honest hour. Longer if the room argues, which is a good sign.
A month of real demand: a ticket export, a call log, or the twenty things people ask for most.
Someone who works the queue, and someone who can change a policy. Not only the person who owns the report.
Three numbers your board has never seen, and the first three sections of your operating model.
Learn the lens Failure demand, flow efficiency, and the arithmetic of queues
Failure demand is demand caused by a failure to do something, or to do it right, for the customer. Value demand is what you actually exist to handle. John Seddon named the distinction in 1989: the point is not to handle failure demand faster, it is to stop causing it. In your world it usually looks like chasing an update, reporting the same fault twice, or asking for something that should never have needed asking.
Flow efficiency is the share of elapsed time that work is actually being worked on. Touch time divided by lead time. Most service and knowledge work runs between 1 and 10 per cent. It is the one number that reliably stops an executive conversation, because every other metric you report, tickets per head, cost per ticket, utilisation, first-line resolution, measures how busy the resource is, and none of them can see how long the work waits.
Little's Law says average lead time equals work in progress divided by throughput. It holds regardless of how work arrives, how long it takes, or what order you do it in. The only requirement is a stable system. So if a queue holds 240 open items and closes 48 a day, average lead time is five days, arithmetically. Not because anyone is slow. You cannot exhort your way out of it: you reduce work in progress, or you raise throughput at the constraint.
Kingman's formula explains why the last few points of utilisation cost so much. The utilisation term is hyperbolic, not linear: at 50 per cent it contributes 1, at 80 per cent it contributes 4, at 90 per cent 9, at 95 per cent 19. Going from 80 to 90 per cent roughly doubles the wait. And utilisation multiplies with variability rather than adding to it, so high utilisation is survivable with steady demand, and catastrophic with spiky demand. The practical consequence is uncomfortable: a contract that prices for ninety per cent utilisation and promises tight response times is internally contradictory.
Where your demand comes from, and where it should sit
Failure demand
Flow efficiency
Paste it in rather than typing it
Chase an update 420 4 26Your demand, in the customer's words
Record what they say, not what your categories call it| Demand | Vol / mo | Type | Share |
|---|
Touch time against elapsed time
Touch = minutes of actual work · Elapsed = hours from raised to resolved| Demand | Vol / mo | Touch (min) | Elapsed (hrs) | Flow eff. | Waiting |
|---|
Leave a row blank if you do not know it. A question mark is more useful than an average.
Score the moment, allocate the mode
1 to 5 · the model places each line in a mode| Demand | Stakes | Emotion | Judgement | Variability | Data vis | Handled today | Mode |
|---|