The counterintuitive Toyota principle that turns lumpy, unpredictable demand into a smooth and profitable operation — whether you make things, deliver projects, run engineers or fill a diary.
Action Briefs are built to help you do something — not just learn about it.
So everything you put into the interactive sections is yours to keep. Work through them, ask for the read on what you’ve put down, and you’ll finish with a real step toward the business — and the life — you actually want.
Levelling came out of a car factory, but the principle has nothing to do with cars. It is about the gap between how unevenly work arrives and how evenly you are able to do it — and that gap exists in every business that delivers anything to anybody. Tap the closest match and the rest of this brief will speak your language.
The Levelling Paradox
Your biggest jobs feel like wins. A major customer rings with something urgent. Your team drops everything, reshuffles the week, works late. It feels responsive. It feels lean. It is actually the most expensive thing you do — and the cost never appears as a line on your accounts.
"Ship to order, but build to a levelled schedule."
This is the levelling paradox. To deliver faster, you have to stop chasing every job. To carry less work-in-progress overall, you sometimes have to hold more of it — in the right place. To be genuinely responsive, you need the discipline to be the tortoise, not the hare.
The Hidden Cost
A 10% wobble in customer demand does not create 10% of disruption inside your business. It amplifies at every hand-off — from the enquiry, to whoever schedules the work, to the people doing it, to your suppliers and subcontractors. By the far end it is a tidal wave. This is the bullwhip effect, and it is quietly eating your margin. Drag the slider and watch it happen.
"The levelled schedule should be considered as the voice of the customer — not the actual customer order, but a defined agreement that represents their needs, smoothed for the benefit of your processes."
The Three Levels
Levelling is not all-or-nothing. Toyota teaches a clear progression — three levels of increasing difficulty. It applies just as cleanly to a diary or a project board as it does to a production line. Start where you are. Move when you are ready.
Getting Started
Toyota commits at roughly 80% of peak demand and covers the gap with planned overtime. The same ratio works whether you are counting units, jobs, appointments or billable days. Do not wait for perfect data — begin with a graph and your own judgement. Once you hold a level for a few weeks, the right number becomes obvious.
Before you can level, the way you work needs basic stability. Not perfection — just consistent enough that a schedule means something. If every week is firefighting and nothing takes the same amount of time twice, start here first.
Choose a level and commit to it. Treat the levelled schedule as the “voice of the customer” — a smoothed agreement about what they need, rather than a live feed of what they happen to be asking for today. Track every deviation and why it happened. That log is where the real learning is.
Know what capacity you actually have, in the same unit you just levelled. Map your people and kit against it. Then find where you need flex — cross-trained staff, a second person who can cover, subcontractors or associates who can take overflow at short notice.
Do not pretend the peaks are not coming. Build a deliberate plan for them: flexible resource, a standing overtime budget, a held slot, a buffer of your most common work ready to go. The goal is to absorb the peak without destabilising everything underneath it.
"The key is to stretch enough to make a great improvement and to challenge your capabilities, but not so much that total failure results."
The Virtuous Cycle
Once your schedule is stable, something changes. Every improvement you make now shows up in the overall result, instead of disappearing into slack somewhere else. No more “pocket improvements” that make one person’s day better and change nothing you can measure. Everything is connected, and every refinement counts.
As you tighten the levelled schedule — smaller batches, faster switching, less work sitting half-done — the weakest links give way. That is not a bug. That is the entire point.
You gather round and fix the thing that broke. Stability returns. Then you tighten again. Each cycle builds real capability: shorter lead times, more flexibility, less work stuck in the system, and a business that competes on something your rivals cannot copy in a quarter.
Do not chase a reduction in half-finished work as the target — whether that is stock on a rack, open jobs on a board, or files sitting at 80% done. Use it as the measure. As the way you work gets more capable, it falls on its own. The real goal is connected flow from enquiry to invoice.
Use it as your scorecard, and watch batch size rather than total. Twenty big batches beats nothing; a hundred small ones of the same total size beats it comfortably — because you can change direction four times as fast for exactly the same amount of work in the system.
Self-Assessment
Five questions, whatever kind of work you do. They place you on the levelling scale and point at your first move. Be honest — this is for you, not for show.
For Complex Businesses
“But every job we do is different” — that is what every business says, and it is almost never true. Toyota’s answer is to group and then rank, and it works because the Pareto principle turns up almost everywhere.
Group your work by what it has in common — the same steps, the same skills, the same kit, the same rough duration. Those groups are your families. Each family gets its own levelled pattern. You are not levelling four hundred different jobs; you are levelling five or six families.
Within each family, rank by how often the work comes up. The top tier gets a slot every day or every week. The second tier at roughly half that frequency. The third at half again. Your most frequent work sets the rhythm, and everything else slots in around it.
This Week
Three things you can do in the next seven days without changing a system, buying anything, or telling a customer no. Between them they give you the two numbers levelling actually runs on.
Plot your output for the last twelve weeks — units, jobs, appointments, billable hours, whatever your unit of work is. Draw a horizontal line at roughly the 80th percentile: the level you clear four weeks in five. That line is your starting level. Everything above it is what you plan to absorb, not what you plan to promise.
Find the work that always arrives late and always causes the reshuffle. Block out one deliberately unbooked slot a week to catch it — a held afternoon, an unallocated engineer, a spare bay. Hold it for four weeks even on the weeks nothing turns up. That empty slot is not waste; it is what stops one surprise wrecking five other jobs.
For one week, tally every time somebody stops one type of work and starts another — a changeover, a site move, a different client file, a room turnaround. Note roughly how long each switch cost. Most owners are shocked by the total. That number is the size of the prize, and it tells you which switch to attack first.
Not just a copy of your answers — your levelling score, what it points to for your kind of operation, and the first peak worth smoothing.
You have the diagnostic and you have the three actions. If you want the read on what your score actually means for your business — and which peak to smooth first — ask for the snapshot. It comes with a fillable worksheet to turn this into a plan.
