Seth's Blog : Ten steps on the road to efficient

Ten steps on the road to efficient

Frederick Taylor taught Henry Ford how to do mass production. Deming brought quality, systems understanding and respect for the worker. And operations research brought insight.

If you have a repeated production process, the method for improving it is almost always the same, regardless of what you and your team produce:

1. Measure before you change. You can’t improve what you haven’t observed. Go to the floor, watch the actual work, time it, and document what’s really happening—not what you assume is happening. Taylor called this time study. operations research calls it data collection. Either way, you start by looking.

2. Map the flow. Trace the path of materials and information from start to finish. Where does work queue up? Where does it sit idle? Where does it move backward? A simple process flow diagram reveals bottlenecks you’d never see otherwise.

3. Identify the constraint. Your system can only move as fast as its slowest step. Find it. Everything else is secondary until you address that bottleneck. (At a buffet, when you double the number of stations of the slowest item, the entire line runs faster.)

4. Separate value from waste. For every step, ask: does this transform the product in a way the customer would pay for? Anything else—waiting, moving, inspecting, reworking—is waste. You don’t need to eliminate all of it, but you need to see it.

5. Standardize the best-known method. This is Taylor’s core insight: once you find a better way, write it down, teach it, and make it the default. Not to control workers, but to create a floor that everyone can build on. Deming’s insight is that variation is the enemy of quality.

6. Reduce variation before you optimize speed. This is Deming’s most important and surprising lesson. A consistent process running at moderate speed beats an erratic one running fast. Get the process under statistical control first.

7. Build in feedback loops, not inspection gates. Smart managers don’t like end-of-line inspection because it’s too late. Instead, give the people doing the work the information and authority to catch problems as they happen. The goal is to make quality intrinsic to the process, not bolt it on after.

8. Optimize the system, not the parts. This is where operations research and Deming converge. Making one station 30% faster can actually make the whole system worse if it just piles up inventory before the next step. Ask: what does this change do to the entire flow?

9. Involve the people doing the work. Taylor got this wrong—he treated workers as interchangeable parts. Deming fixed this: the people on the floor know things management never will. Create structured ways to capture that knowledge. Invest in reducing fear so people will share what they know.

10. Iterate in small cycles. Plan-Do-Study-Act is Deming’s learning wheel. Don’t redesign everything at once. Make a small change, measure the result, learn from it, adjust. Then do it again. The factory you want isn’t built in a single leap—it emerges from dozens of small, informed improvements compounding over time.

The meta-principle underneath all ten: respect the system and the people in it. Change the system before you blame the people.

And don’t get efficient at doing something you’d rather not be doing at all.