You’ve probably heard of Eliyahu M. Goldratt. If you’ve spent any time in a manufacturing plant or a high-stakes corporate boardroom, his name carries a certain weight, almost like a secular prophet of efficiency. His 1984 business novel, The Goal, didn't just sell millions of copies; it fundamentally shifted how we think about work. But here’s the thing. Most people treat the goal a process of ongoing improvement as a buzzword or a one-time project. They think they can "optimize" a department and then go back to business as usual.
That’s a mistake. A massive one.
Goldratt wasn't just talking about fixing a broken machine on a factory floor in a fictional town. He was describing a total shift in mindset. It’s about the "Theory of Constraints" (TOC). It sounds academic, but it’s actually pretty gritty. It’s about finding the one thing—the bottleneck—that’s holding everything else back. If you aren't focusing on that specific constraint, you’re basically just making noise. You're "optimizing" things that don't matter, which is a fancy way of wasting time.
Honestly, it’s frustrating to watch. I see companies spending millions on new software to speed up a process that isn't even the bottleneck. That’s like putting a turbocharger on a car that has no wheels. It makes no sense, yet we do it every day because it feels like "improvement."
The Myth of Local Optimums
Most managers are obsessed with "busy-ness." They want every person and every machine running at 100% capacity. If a worker is sitting idle, the manager panics. They find something—anything—for that person to do.
This is what Goldratt called a "local optimum." And it's a trap.
Think about it this way. If you have a production line where Step A is super fast but Step B is slow, making Step A run at full speed just creates a giant pile of inventory in front of Step B. You haven't helped the company make more money. You've just tied up cash in a pile of half-finished parts. In the context of the goal a process of ongoing improvement, true efficiency is only measured at the very end of the line.
Everything else is just secondary.
✨ Don't miss: Where’s My Refund? Why Your Tax Money Is Taking Forever This Year
Alex Rogo, the protagonist of the book, figures this out during a Scout hike. He notices the line of hikers keeps stretching out. The fastest kids are way ahead, and the slowest kid, Herbie, is falling behind. To fix the "process," Alex doesn't tell the fast kids to run faster. That would just make the line longer. Instead, he puts Herbie at the front and distributes the heavy gear from Herbie’s pack among the other kids. The whole group starts moving faster as a single unit.
That is the essence of TOC. You manage the constraint, not the individuals.
The Five Focusing Steps That Actually Work
If you want to implement the goal a process of ongoing improvement, you can't just wing it. There’s a specific sequence. Goldratt laid out five steps, and they are surprisingly brutal in their simplicity.
First, you Identify the constraint. This is harder than it looks. Sometimes the bottleneck isn't a machine; it’s a policy. Or a person who insists on approving every single email. You have to find the "Herbie" of your organization.
Second, you Exploit the constraint. This doesn't mean "buy a new machine." It means make sure the current constraint is never idling. If your bottleneck is a specific piece of software, make sure someone is always operating it, even through lunch breaks. Don't let your constraint waste time on "junk" work.
Third, you Subordinate everything else. This is where most people fail. It means everyone who isn't the bottleneck needs to slow down or change their rhythm to match the bottleneck. It feels counter-intuitive. It feels like being lazy. But if you don't subordinate, you’re just creating piles of "work-in-progress" that clog up the system.
Fourth, you Elevate the constraint. Okay, now you can buy the new machine or hire the extra person. You only spend money to increase capacity after you’ve squeezed every drop of efficiency out of what you already have.
Fifth, Don’t let inertia set in. Once you elevate the constraint, it will likely stop being the bottleneck. Something else will become the new "Herbie." You have to go back to step one immediately. It’s a loop. It never ends. That’s why it’s called ongoing improvement.
Throughput, Inventory, and Operating Expense
We need to talk about the math, but don't worry, it’s not complex. Goldratt hated traditional cost accounting. He thought it was misleading and often led to bad decisions. Instead, he proposed three measurements:
- Throughput: The rate at which the system generates money through sales. If you didn't sell it, it’s not throughput.
- Inventory: All the money that the system has invested in purchasing things which it intends to sell.
- Operating Expense: All the money the system spends in order to turn inventory into throughput.
In the world of the goal a process of ongoing improvement, the objective is simple: Increase throughput while simultaneously decreasing inventory and operating expense.
Most companies focus heavily on cutting operating expenses (layoffs, cheap supplies). But Goldratt argued that throughput is the most important because it has no theoretical limit. You can only cut expenses so far before you have no company left. But you can increase throughput indefinitely.
I remember a real-world example involving a gear manufacturer in the midwest. They were struggling with lead times. They thought they needed a bigger warehouse. After looking at their process through the lens of The Goal, they realized their "Herbie" was actually the heat-treat oven. They were running small batches of parts through the oven just to keep the "efficiency" numbers up on the machines before it. By stopping those machines and only running what the oven could handle, their lead times dropped from weeks to days. They didn't need a warehouse; they needed a better schedule.
The Psychological Barrier to Change
Why is this so hard? Honestly, it’s ego.
Managers want to feel like they are "optimizing" their specific department. If a VP of Manufacturing sees a machine standing still, they feel like they’re failing. They’ve been trained for decades that "Idle time = Waste."
But in a TOC environment, idle time for a non-constraint is actually a virtue. It means the system is balanced. It’s a hard pill to swallow. You have to be okay with people "doing nothing" if their work would only create a bottleneck elsewhere.
We also have to deal with the "we've always done it this way" syndrome. This is the ultimate enemy of the goal a process of ongoing improvement. People get comfortable with their silos. They like their piles of inventory because it feels like a safety net. In reality, inventory hides problems. It’s like a high water level in a river; it covers up the rocks at the bottom. When you lower the inventory (the water), you finally see the rocks (the broken processes) and can actually fix them.
Real-World Limitations and Criticisms
Is TOC perfect? No. Nothing is.
Critics often point out that Goldratt’s model assumes a relatively stable environment. In a world of extreme volatility—like the 2020 supply chain collapses—having "zero inventory" can be a death sentence. We’ve seen a shift lately toward "Just-in-Case" instead of "Just-in-Time."
However, even with "Just-in-Case," the core logic of the bottleneck still applies. You still need to know where your constraint is. Even if you decide to hold more inventory to buffer against a global pandemic, you should still be managing your internal flow based on the five focusing steps.
Another critique is that TOC is very "factory-centric." It’s easy to see a bottleneck on a production line. It’s much harder to see a bottleneck in a software development team or a marketing agency. But it’s there. Usually, it’s a specific "Subject Matter Expert" who has to approve everything, or a QA process that’s understaffed. The "Herbie" just wears a hoodie instead of a hard hat.
How to Start Improving Today
You don't need a consultant to start applying the goal a process of ongoing improvement. You just need to look at your work differently.
- Stop looking at individuals and start looking at the flow. Where does work "pile up"? Is there a specific person's inbox where things go to die? That’s your constraint.
- Stop rewarding "busy-ness." Reward throughput. If someone finishes their work and has nothing else to do that helps the bottleneck, let them read a book or learn a new skill. Don't force them to create "busy work" that just clutters the system.
- Focus on the "Drum-Buffer-Rope" method. The "Drum" is the constraint (it sets the beat for the whole plant). The "Buffer" is the pile of work in front of the constraint to ensure it never runs out. The "Rope" is the communication back to the beginning of the line to only release new work when the "Drum" has processed something.
- Check your policies. Often, the biggest bottleneck is a rule that no longer makes sense. "We have to use this vendor because they are 5% cheaper," even though that vendor is three weeks late every time. That 5% savings is costing you a fortune in lost throughput.
Ongoing improvement isn't a destination. It’s not a plaque you hang on the wall. It’s a constant, sometimes annoying, process of hunting for the next thing that’s slowing you down. It requires humility because you have to admit that the way you’re doing things now—the way that maybe even made you successful—is exactly what’s holding you back from the next level.
The goal isn't just to be better; it's to stay better by never stopping the search for the next Herbie.
Actionable Next Steps:
- Identify your current Herbie: Look for the largest pile of unfinished work in your current project or department. That is your constraint.
- Protect the constraint: Ensure that the person or machine identified as the constraint is not doing tasks that someone else could do. Strip away their administrative burden.
- Communicate the "Why": Explain to the "fast" members of the team why it is okay (and necessary) for them to slow down to match the pace of the constraint.
- Audit one "Old Rule": Find one company policy that everyone complains about and trace whether it actually helps or hinders throughput. If it hinders, kill it.