The Five Focusing Steps

You now know the goal, how to measure progress, and that a constraint sets the pace of the whole system. The Five Focusing Steps turn that into a process of ongoing improvement.

1. Identify the system’s constraint.
2. Decide how to exploit the constraint.
3. Subordinate everything else to that decision.
4. Elevate the constraint.
5. If a constraint has been broken in a previous step, go back to step 1, and don’t let inertia become the constraint.

Feel it first: a water pipe

Water flows through a pipe made of five sections. The thinnest section limits the flow for the whole pipe. Widen a section by tapping it and watch what happens to the flow, and to where the limit sits.

Notice three things. Widening any section other than the thinnest does nothing. Widening the thinnest one helps, but only until another section becomes the thinnest. And you could only see that by going back to the start and looking again. That’s the five steps in miniature.

Step 1: Identify the constraint

You can’t manage what you haven’t found. Signs to look for: the biggest queue of work waiting in front of it, the resource everyone is chasing, the place where overtime is always requested. If you compare load against capacity, the constraint is the resource whose load is closest to, or above, what it can deliver. If no resource is overloaded and there’s still unsold capacity, the constraint is in the market.

Step 2: Exploit the constraint

Exploit means: get the most out of the constraint you already have, before spending money on more. Because an hour lost at the constraint is an hour lost for the whole system, and an hour saved anywhere else is a mirage.

Typical ways to exploit:

• Never let the constraint stand idle: cover breaks and shift changes, keep a small buffer of work in front of it.
• Don’t waste its time on defective parts: inspect before the constraint, not after.
• Don’t spend its time on work that doesn’t bring throughput, and give priority to the work that earns the most per hour of the constraint.

Goldratt observed that often, simply by not wasting what was available, a huge constraint turned out to have plenty of capacity hiding in it.

Step 3: Subordinate everything else

This is the hardest step, because it asks the non-constraints to behave in a way that looks wrong by the old measures. A non-constraint should work at the pace of the constraint, not at its own maximum. If it runs flat out, it only builds piles of work in progress: more Investment, longer lead times, and no extra sales.

“When there is nothing to do, the best thing to do is nothing.”

That line often gets a laugh in class, until people see their own shop floor in it. Subordinating can mean: releasing material only as fast as the constraint consumes it, maintenance answering the constraint first, purchasing and planning protecting its schedule, and sales selling the products that make the most of its time.

Step 4: Elevate the constraint

Only now, after exploiting and subordinating, do you consider adding capacity: a second shift, another machine, outsourcing, hiring. Note the word evaluate in many versions of this step: compare the alternatives on what they do to T, I and OE. Elevating often costs money, so do it after you have squeezed out what the first three steps give you for free.

When you elevate, keep protective capacity on the non-constraints: a non-constraint running at 100% has no slack to recover from a disruption and starts to behave like a second constraint.

Step 5: Go back to step 1, and beware of inertia

When you elevate a constraint enough, it stops being the constraint. Something else now limits the system, and the policies you built around the old constraint may now be in the way. That’s why the step warns against inertia: rules that were right last year can become today’s policy constraint.

Put it to work: a furniture workshop

A small workshop makes wooden chairs in five steps. Each chair sells for $100 and uses $40 of material, so each extra chair sold adds $60 of throughput. Customers want 120 chairs a day. Work through the five steps.

How to tell if you’re managing the constraint well

• Customers get what they were promised, on time.
• The constraint rarely waits for work, and rarely works on the wrong thing.
• Work in progress and inventory are lower than before, and so are lead times.
• Improvement efforts are aimed at the constraint, and people know where it is.
• Cash flow gets better, because money spends less time tied up in material.

Check your understanding

Answer all five. Get four right to complete the lesson.

1. What does “exploit the constraint” mean?

• Buy more of it
• Get the most out of the constraint you already have
• Work its operators harder
• Move it to another department

Stop wasting its time first. It’s usually the cheapest and fastest gain.

2. A non-constraint can make 160 units a day, and the constraint 100. Subordinating means the non-constraint should make about…

• 160, to keep efficiency high
• 130, as a compromise
• 200, with overtime
• 100, the pace of the constraint, plus whatever the buffer needs

Anything more turns into work in progress, not sales.

3. Why should you exploit and subordinate before you elevate?

• Because elevating usually costs money, and the first steps often release capacity for free
• Because elevating is never needed
• Because the order doesn’t matter
• Because accounting requires it

Many “constraints” have a lot of hidden capacity, wasted on idle time, defects and the wrong work.

4. What is protective capacity?

• Capacity reserved for the constraint’s operators
• Spare capacity at the constraint
• Spare capacity on non-constraints, so they can recover from disruptions without starving the constraint
• Insurance against fire

Without it, a non-constraint at 100% load starts behaving like a second constraint.

5. Step 5 warns against “inertia”. What does that mean here?

• Machines slowing down over time
• Keeping the rules made for the old constraint after the constraint has moved
• People resisting overtime
• Customers not changing their orders

Yesterday’s smart rule can become today’s policy constraint.