Lean Six Sigma for manufacturers: where it pays off, and where it does not

Lean Six Sigma usually gets sold as a belt program and a wall of certificates. After three decades running plants, I have watched it deliver real money, and I have watched it become a training budget that produced nothing. The difference is rarely the tools. It comes down to whether the work was aimed at a problem that actually mattered.

Two ideas, one goal

Lean and Six Sigma started in different places. Lean came out of Toyota and goes after waste: motion, waiting, overproduction, anything the customer would not pay for. Six Sigma came out of Motorola and goes after variation, the scatter that makes a good process produce bad parts. In practice you need both, because a fast process that is unpredictable still ships defects, while a precise process buried in waste still loses money. So the goal is simple. Make the work flow, and make it consistent.

Start with the problem, not the belt

The most common mistake I see is backwards sequencing. A plant trains a dozen Green Belts, then goes looking for projects to justify the training. That order almost never works. The projects that pay off start the other way around. First you find the line that is bleeding, whether that is scrap, downtime, or a customer who keeps rejecting shipments, and only then do you bring the right tool to it. Sometimes that tool is a full DMAIC project. Just as often it is a single afternoon of value stream mapping and a few changes anyone could have made, if someone had simply looked.

DMAIC, briefly and honestly

DMAIC is the backbone: Define, Measure, Analyze, Improve, Control. The phases that get skipped are the first and the last. Teams rush past Define and end up solving a symptom. Then they declare victory at Improve and never build the Control step that protects the gain. As a result, the metric drifts back within a quarter, and everyone concludes the program does not work. The program was fine. The discipline was missing.

What it looks like when it works

A real engagement is unglamorous. You watch the line. You measure what is actually happening instead of what the standard says should happen. You find the one or two variables driving most of the loss, you change them, and then you put something in place, a control plan, a check, a chart, so the change holds after the team moves on. Done that way, the savings are real and they show up where the finance team can see them. The certificates are a nice byproduct, not the point.

The bottom line

Lean Six Sigma is not a religion, and it is not a filing cabinet of belts. It is a way of aiming attention at the few things costing you the most, and then making the fix stick. So if your last improvement effort faded, the tools probably were not the problem. Start with the bleeding, finish with the control, and the rest tends to follow.