Lean – value, waste and the system Toyota built

This is the script, written before recording. The text below is what gets said, and the lines in italics are production cues – they are not read aloud. Once the episode is recorded, this text is replaced by what was actually said.

Length: allow 18–20 minutes. This is the densest episode so far.

If it runs long: cut “The toolbox” down to kanban, jidoka and gemba, and say the rest is in the article. Do not cut “Muda, mura and muri” or the criticism – those two are what make this better than a standard lean introduction.

The opening

(Start with the misuse of the word. It is the best way in, and it is documented in the article.)

“We need to get leaner.”

When people say that, they usually mean they are going to be fewer. That someone will do the same work with less staff.

That is roughly the opposite of what the system actually says.

Today we are going to look at what lean really is – where it comes from, what it actually prescribes, and what criticism it has rightly drawn.

Where it comes from

(Graphic: timeline or a simple card – Toyota, Ohno, Krafcik, MIT.)

Lean is a Western umbrella term for the Toyota Production System, developed at Toyota in Japan in the decades after the Second World War. Taiichi Ohno is regarded as the chief architect, with Shigeo Shingo as a key contributor.

(This is the key to understanding the whole system. Say it clearly.)

And the starting point was scarcity.

Japanese industry had little capital, little space and few raw materials. They could not copy American mass production with its large inventories and long runs. They could not afford to.

So the system that emerged was an answer to scarcity. And that explains why it puts so much weight on not making anything before it is needed.

The word “lean” itself came much later, and not from Toyota. John Krafcik used it in an article in 1988, while he was attached to the IMVP research programme at MIT. Two years later came the book The Machine That Changed the World, which carried the term far beyond the car industry.

Value, and everything else

(Graphic: the definition of value and the test.)

The whole system rests on one distinction.

Value is what the customer is actually willing to pay for. Everything else is waste.

And the test is simple: would the customer pay more for this step if they saw it?

Welding the frame – yes. The part sitting in stock for three weeks first? Being moved back and forth between two buildings? Someone filling in a form nobody reads? No.

(Important qualification. Do not skip this one.)

But note the wording. Waste is not the same as unnecessary work.

Much of the waste in an organisation is work that has to be done the way the system is set up now. The point is to change the system – not to ask people to run faster.

That is worth carrying with you, because it is exactly where most lean programmes go wrong.

The seven wastes

(Graphic: the table with a factory column and an office column. That translation is what makes the episode relevant to everyone.)

Ohno set out seven types of waste. They were written for a factory, but they translate to office work surprisingly directly.

Overproduction is making more than the next step needs right now. In an office: reports nobody reads, meetings nobody needed.

Waiting. The machine stands idle and the operator waits for parts. In an office: you wait for an approval, an answer, an access right.

Transport. Parts moved between buildings and halls. Or: the case is passed between four departments.

Over-processing. A finer tolerance than the customer needs. Or fifteen fields where five would do.

Inventory. Halls full of half-finished goods – or queues of unfinished cases and unanswered enquiries.

Motion. The operator reaching for a tool. Or you searching four systems to find one number.

Defects. Scrapped parts and rework. Or bad data that has to be corrected further down the line.

(This is the key point in this section.)

Ohno considered overproduction the worst of them, and the reasoning is worth noting: it creates all the others.

Make more than is needed and you get inventory. You get transport. You get waiting. And you get defects that are not discovered until long after they happened.

Many have since added an eighth – unused talent: that the people who know the job best are not asked. It was not on Ohno’s list, and you should be honest about that. But it points at something the system does assume, namely that the improvements come from the people doing the work.

The five principles

(Graphic: the five, numbered.)

In Lean Thinking from 1996, Womack and Jones set out five principles. They have become the most common way into lean outside the car industry.

Specify value – seen from the customer, not from your own department.

Map the value stream – every step from order to delivered. Including the ones nobody has ever drawn.

Create flow – get the work moving without stopping in queues.

Let the customer pull – nothing is made before the next step actually asks for it.

And seek perfection – not as an end point, but as never declaring it finished.

(The point of the order.)

The order is not decoration. Create flow before you know what the customer values, and the result is that the wrong thing gets produced faster.

And that is a more expensive mistake than producing it slowly.

The first step, incidentally, is the one most often skipped.

Pull instead of push

(Graphic: push versus pull, side by side. This is the biggest practical difference.)

Now to the biggest practical difference from traditional planning.

Push means the plan decides. Production follows a forecast, finished goods are pushed on to the next step, and if the forecast is wrong, inventory builds up. That is classic MRP and ERP logic.

Pull means demand decides. Production happens when the next step asks for it. A signal triggers replenishment – and that is exactly what kanban is. Inventory stays low because it is not refilled unprompted.

(The good story. Take it.)

Ohno has said the idea came from American self-service supermarkets: the shelf is refilled when someone has taken something, not according to a plan made in advance.

But there is one precondition that is easy to miss: it requires short changeover times. Without them, small batches become too expensive, and the whole pull idea collapses.

(The ERP link – relevant to your audience.)

And this is where lean and a traditional ERP setup pull in different directions. Material planning in ERP is push by default: requirements are calculated from forecasts and bills of materials, and orders are released according to plan.

That does not rule out lean. But it does mean the two have to be set up deliberately in relation to each other – otherwise you end up planning one way and running the floor another.

Muda, mura and muri

(Graphic: the three Ms. This is the most useful part of the episode – do not cut it.)

And here Western lean parts ways with Toyota.

Most people know muda – waste. The seven we just went through. It is the only one most people look for.

But there are two more.

Mura is unevenness. Quiet Monday, chaos Friday. Swings in load and staffing.

Muri is overburden. People and machines run beyond their capacity, over time.

(The point about how they connect.)

And the three are connected: unevenness creates overburden at the peaks and waste in the troughs.

Hunt only for waste without doing anything about the swings, and you get a system that looks efficient on paper – and breaks in a busy week.

Toyota’s answer to mura is called heijunka: levelling.

The toolbox

(Graphic: the glossary. Cut to kanban, jidoka and gemba if the episode runs long.)

The Japanese words are one reason lean seems more mysterious than it is. Most of them mean something quite ordinary.

Kaizen means improvement. Many small improvements in daily work, made by the people doing it.

Kanban means sign or signal. A card or an empty bin that says “refill”. The engine of pull.

Jidoka is automation with a human touch. Equipment that stops itself when something is wrong – and the right to stop the line.

Poka-yoke is mistake-proofing. Design that makes assembling it wrongly physically impossible.

Heijunka is levelling. Gemba is the real place – go where the work happens and see for yourself, rather than reading a report. And 5S is five words that all begin with s: sort, set in order, shine, standardise and sustain.

(The jidoka point is the best thing in this section.)

Jidoka is the one most often left out when lean is introduced, and the one that says most about the culture. An operator who sees something wrong has the right to stop the entire line.

It is cheaper to stop for one minute than to make two hundred parts with the same defect.

The criticism

(Four objections. This is what separates the episode from a brochure.)

And now the criticism, because it is real.

The word is used as cover for cutting staff. Cost-cutting exercises that have nothing to do with the Toyota system, where the result is fewer people doing the same work. That is muri – the opposite of what the system prescribes. But the label sticks anyway.

Work intensity. Even where lean is introduced by the book, critics have pointed out that production without slack can mean a higher and more uniform pace, and less breathing room. It has been raised both in industry and in healthcare, and it deserves to be taken seriously rather than dismissed as a misunderstanding.

Supply chain fragility. And this one has gained the most weight in recent years.

(Say this sentence slowly – it is the sharpest in the episode.)

Just-in-time is efficient and fragile for exactly the same reason: there is no inventory to fall back on.

That is not a flaw in the method. It is a cost of it. Low inventories free up capital and expose problems quickly – and those same low inventories mean one stoppage anywhere propagates through the whole chain.

(The link back to the bullwhip episode.)

If you saw the episode on the bullwhip effect, you recognise the mechanism. Remove the inventory without sharing information upward, and you have removed the shock absorber and kept the bumps.

And finally: tools without culture. The most common reason programmes stall. Boards and kanban cards are the visible part, and the easy part to copy. The hard part – that management actually goes out into the operation, that an operator dares to stop the line, that suggestions from the floor get implemented – is what decides whether the rest means anything.

Introduce the tools alone, and they tend to be gone again within a year or two.

What holds up

(Graphic: the four points. Calm ending.)

With the decoration stripped away, there are a few ideas left that hold up well outside a car factory.

That it pays to look at the flow through the whole chain rather than optimising each department separately. A department can become more efficient and make the whole slower.

That queues of unfinished work are a problem in themselves – whether that is parts in a hall or cases in an inbox.

That improvement should be a habit rather than a project.

And that the people doing the work usually know best where it grates.

(Final line. Let it stand alone.)

The rest – the words, the boards, the certifications – is packaging.

Useful packaging. But packaging.

(End card.)