The Constraint You Have Is Not Always the One You Want

Hello TOC Flow Community!

Last month I told you to exploit before you elevate. Southwest found the equivalent of sixteen aircraft in five minutes. There’s usually capacity already there, more than you think.

But there’s something else I want to share.

There's a question that sits underneath the whole exploit conversation, and it changes how much effort you should pour in.

The Question Asked Less Often

Where should your constraint be? Not where is it. Where should it be.

Most operations never ask this. The constraint is wherever it happens to be, usually by accident.

A machine somebody bought fifteen years ago.

A department that never got staffed properly.

A step in the process that everyone works around.

That's an accidental constraint. The system landed there on its own. Nobody decided it.

And here's why the question matters.

If your constraint is sitting somewhere you don't want it long term, there's a limit to how much effort you should pour into exploiting it. We know Exploiting works. We spent all of last month on that. But if you're about to move the constraint anyway, every hour of deep exploitation work you invest there has a short payback window.

The investment never gets time to earn out.

You exploit much deeper on a constraint you intend to keep than on one you intend to move.

Where the Constraint Should Live

So how do you decide? A few things I look for:

Where is capacity most expensive to add? 

If a resource costs millions and takes eighteen months to duplicate, that's a strong candidate for a permanent constraint. If a resource costs forty thousand dollars and takes six weeks, it has no business being your constraint.

Can you protect it? A good strategic constraint is one you can buffer, feed reliably, and keep loaded. If your constraint wanders week to week, you can't build a system around it.

Does it sit where your advantage lives? The best constraints are the ones your competitors also have and can't easily solve.

A Look at Semiconductors

The clearest example of a deliberately kept constraint I can point to is advanced chip manufacturing.

Lithography is the constraint in a leading-edge fab, and everybody in the industry knows it. The EUV scanners that pattern the smallest features on a chip run somewhere between $150 and $200 million each for the current generation. The newest machines are reported to exceed $350 million. They come from a single supplier. Lead times are measured in years. Installing one takes months.

Nobody is elevating that constraint casually.

So watch what fabs do instead. They exploit it to a degree that would look obsessive anywhere else. Tool uptime is tracked relentlessly. Wafers are staged so a scanner never waits. Maintenance is scheduled around exposure windows. Process engineers work to reduce the number of layers that need the expensive tool at all. Entire process technologies get designed around how many passes through that machine they require.

That is exploit taken to its limit. And it makes sense, because that constraint is not going anywhere. The depth of the exploitation matches the permanence of the constraint.

Now imagine a fab that spent that same intensity optimizing a piece of $2 million support equipment. Same rigor, same engineering hours, applied to a resource they could simply buy another of. That would be a waste of the smartest people in the building.

Same technique in both cases. Knowing which constraint deserves it is the difference.

A quick reminder

Everything from last month still holds. Even on a constraint you plan to move, the fast, cheap exploit comes first. Protect it from starvation. Stop putting the wrong work in front of it. Offload what doesn't belong there.

Two reasons:

1. You'll frequently discover you need far less new capacity than you thought, and sometimes none.

2. And it's a good way to size an elevation. If you buy capacity before you exploit, you're buying based on a number that includes your own waste.

Exploiting is never a mistake. The mistake is applying the same depth of effort everywhere, regardless of whether the constraint has a future.

Your September Challenge

This month, take your current constraint and answer one question.

"Is this where I want my constraint to be three years from now?"

If yes, go much deeper than you have. Most companies stop exploiting far too early on the constraints that matter most.

Ask what else you could offload from it.

Ask what work reaches it that shouldn't.

Ask whether it ever waits.

…The worst outcome is spending three years perfecting a constraint that never should have been your constraint at all.

We hold Flow Workshops around the country. Find one near you today! If you don't see your city listed, reach out.