High-Level Thoughts

Very much a “zero to one” book, where after you read it you see the world differently. Extremely helpful meta-tactic for evaluating how to improve systems, and I wish I had read it sooner.

Summary Notes

My Executive Summary

The three most important areas of focus:

  • Throughput: the rate the system generates money through sales. Money coming in.
  • Inventory: all the money invested in things it intends to sell. Money in the system.
  • Operational Expense: The money spent to turn inventory into throughput. Money you pay to make throughput happen.

The Goal: Increase throughput while simultaneously reducing both inventory and operating expense.

The system only moves as fast as the bottleneck, so find the bottleneck, see what has a big pile of work in front of it.

Five Focusing Steps for Continuous Improvement:

  1. Find the bottleneck.
  2. Get more from the bottleneck’s current resources.
  3. Redirect existing resources towards the bottleneck, don’t produce more than it can or you’ll build up inventory.
  4. Increase resources for the bottleneck ONLY once you’ve squeezed the most you can out of it.
  5. Find the new constraint and return to step 1, consider increasing resources to other areas to match the bottleneck’s improvement.

All Notes

You could think you’re running an efficient system, but your thinking might be wrong. If you didn’t increase sales, throughput, or decrease costs, you didn’t increase productivity.

If you don’t know what the real goal is, which you could very well be wrong about, then you can’t figure out what to do to reach the goal. And the goal of any business is to make money.

Keeping people working and making money aren’t the same thing. Just because you’re paying for someone doesn’t mean they should be busy all the time, it could be harmful.

The Goal: Increase net profit while increasing return on investment and increasing cash flow

Those money making measurements are difficult to use day to day though, so you can use:

  • Throughput: the rate at which the system generates money through sales. You can also think of it as money coming in.
  • Inventory: all of the money that the system has invested in purchasing things which it intends to sell. Money that’s currently in the system.
  • Operational Expense: All the money that the system spends in order to turn inventory into throughput. Money that you have to pay out to make throughput happen.

The Goal Reframed: Increase throughput while simultaneously reducing both inventory and operating expense. Not to do them in isolation, but to do them all together.

If you keep everyone and everything working at full capacity, you’ll naturally build up inventory by creating excess work. A plant where everyone is working all the time is very inefficient. You can’t have a “balanced plant” without doing excess work.

Because of dependent events and statistical fluctuations, you’ll naturally run into bottlenecks in the system, kind of like the fattest kid on a hike slowing everyone down. The whole system only moves as fast as the bottleneck, so it makes sense to focus on increasing the bottleneck’s capacity, and tying the rest of the system’s rate to the rate of the bottleneck. In the analogy, put the fat kid at the front of the line, and make his backpack as light as possible so he can walk faster.

Since output can only deviate up to the maximum level determined by its dependent events, but it can deviate down much lower, with successive dependent events you’ll get further and further negative fluctuations. A kid can come to a halt, but he can only catch up as much as the kid in front of him, he can never catch up past the point the previous kid has already walked.

Bottleneck: “A bottleneck is any resource whose capacity is equal to or less than the demand placed upon it. And a non-bottleneck is any resource whose capacity is greater than the demand placed on it.” [139]

An easy way to find a bottleneck is to look and see what has a big pile of work in front of it [144]

To optimize the system, make the flow through the bottleneck equal to the demand from the market. Or, a tiny bit less than the demand from the market. [139]

In most cases you’ll have capacity that is hidden from you because some of your thinking is incorrect. The first thing you should always do is see exactly how the bottlenecks are currently operating, and if you can change how they’re used to increase their capacity (before simply hiring or buying more tools). [152]

If you lose even one hour on the bottleneck, you have lost it forever. You can’t get it back somewhere else in the system. Your throughput for the entire system will be lower by whatever amount the bottleneck produces in that time. [153]

Lost time on the bottleneck is lost throughput which means you’ve lost the total output of the whole system. If your whole plant earns $1,000 an hour, then an hour lost on the bottleneck is $1,000 lost. Make sure it’s time isn’t wasted by: [157 / 159]

  • Sitting idle and not getting used during some break
  • For it to be spending time on output that’s already defective
  • Making it work on something you don’t need

You can increase the capacity of the bottleneck by: [159]

  • Only making it work on what will increase throughput today
  • Taking some of the load off the bottleneck and giving it to non-bottlenecks

When you make a non-bottleneck do more work than the bottleneck, you create excess inventory and thus lose money. [210]

What you have to do is figure out how to release materials at the start of the process exactly according to the capacity of the bottleneck. [216]

If you reduce your batch size, you increase throughput by reducing inventory held and reducing the amount of cash tied up at any one time. It also lets you move faster, since the gaps will be smaller since the time to process a batch will be lowered as well, and your total lead time on any project condenses. [232]

Three simple questions: What to change, what to change to, and how to cause the change. [337]

Five Focusing Steps:

  1. IDENTIFY the system’s constraint.
  2. Decide how to EXPLOIT the system’s constraint.
  3. SUBORDINATE everything else to the above decisions.
  4. ELEVATE the system’s constraint.
  5. If in the previous steps a constraint has been broken, go back to step 1, but do not allow inertia to cause a system constraint.

Enjoyed this? Be sure to subscribe!