Theory of Constraints for Beginners: Identify Bottlenecks and Improve System Efficiency

1. Manage constraints in your business operations: Operational challenges such as insufficient capacity, late customer deliveries, and rising inventory levels affect the organizations across the industry, whether in manufacturing, healthcare, customer service, operations, airport, or even content creation. Performance issue often traced back to one critical factor, a constraint within the system. Understanding what limits your operations and knowing where to focus improvement efforts can significantly enhance capacity, flow, and overall performance. Welcome to this course on constraint and bottleneck management. Hi, I'm a consultant and coach specialized in operational excellence and business performance improvement. In this class, you will learn how to identify bottlenecks, understand systems, constraints, and apply practical methods to improve throughput and reduce delays. Operational systems are the backbone of most organizations. Every business operates as a sequence of interconnected process, and as the saying goes, a chain is only as strong as its weakest link. Constraint management provides a structured approach to identifying the true limiting factors in a process. Prioritizing improvement efforts effectively increases output without unnecessary investment. Create sustainable performance gains. By the end of this class, you will be able to analyze your own processes and pinpoint where improvement will have the greatest impact. If you are ready to take a more focused strategic approach to operational excellence, let's begin. I will see you inside the class. 2. Everyday bottlenecks and system efficiency: Lesson two. A weekend. You may have left the office, the factory floor, or the customer service desk behind. But operational systems and constraints do not disappear. They are everywhere. Think about a typical Saturday. You decide to drive out of the city, everything moves smoothly until you reach that one major junction where four lane merges into two, traffic slows down. Your car stucks up. Nothing is wrong with the road itself. The constraint is simply the merging point. Or imagine ordering a coffee at a Psy cafe. There are three people taking the orders, two preparing the drink, but only one at the payment terminal. The line does not move because of demand. It moves at the speed of this single payment point. That is constraint management in action. Now, let's take a larger example. A day at an amusement park. You park the car, you queue for the ticket, you wait to enter. You navigate to the roller coaster. The ride itself lasts only 1 minute, and it fits ten people. On the paper, that sounds like 600 people per hour. But the queue moves far slower. Why? Because the real constraints may not be the right duration. It might be how long it takes to load and unload the passengers. It might be the safety check. It might be the staff coordination. The slowest step determines the system's output. The same principle applies at a food stall. People hesitate while choosing. Someone struggles with the payment. Another blocks the counter while adding some chocolates. The grill might be fast, the bottleneck might be the decision making point or the payment. Constraint management is not about working harder. It's about identifying what truly limits the performance. You see at a hospital where the operating rooms are available, but discharge process delays new admissions. At the airport, where the security screening limits the passenger flow. In a call center where the agents are available, but the system login time slows the response. A household where everyone is ready to leave, but one person cannot find their keys. In every case, the system moves at the pace of its constraint. The key insight is simple. The improving parts of the system that are not the constraint does not increase overall performance. If traffic is stuck at two lane merges, widening the road elsewhere will not help. If checkout is the bottleneck, adding more sales staff on the floor will not reduce the queue. Constraint management teaches you something important. Identify the true limiting factor, focus improvements where it matters the most. Increase the throughput without unnecessary investment. Improve flow instead of creating local efficiencies. Whether in business or public service or everyday life, system behaves the same way. The question is not whether a constraint exists. The question is, can you see it? And once you can see it, do you know how to manage it? That is what we will learn in this course. 3. When You Are the “Job” Moving Through the System: Lesson three, when you are the job moving through the system. Sometimes the easiest way to understand constrained management is to stop thinking like a manager and start thinking like an item being processed because often that item is you when you sit in traffic at a busy intersection, you feel the slowdown. When you wait for a table at a restaurant, you feel the delay. When you stand in a line at an amusement park ride, watching it operate again and again, while the queue barely moves, you feel the friction. In those moments, you are the material moving through the operational system. And the system is processing you. These everyday experiences are not random inconveniences. They are structured systems with inputs, process, and outputs. The traffic network process vehicles per hour. The restaurant processes guest per hour. The amusement park ride processes people per cycle, just like the logistic company processes packages. Or a team processes task, or a hospital processes patient, or a marketing agency produces campaigns per week. Whether the system moves material, people, task, or data, the principle is the same. The system's output is determined by its constraint. If the restaurant has 20 empty tables, but only one chef, the chef is the constraint. If the hospital has available doctors but no available ICU beds, the beds are the constraint. If a call center has many agents, but a slow CRM system, the software is the constraint. If a factory has fast machines, but a frequent breakdown at one station, that station limits the entire line. The constraint is the weakest link, but it is also the most important link. It determines throughput, it determines the lead time, it determines the revenue and the customer satisfaction. Improving non constraints may create local efficiencies, but it does not increase the total system output. If the checkout is slow, adding more sales assistance on the floor does not reduce the queue. If the security screening is slow at the airport, adding faster boarding procedures does not increase the passenger flow. The system only moves as fast as its limiting factor. The goal is to simplify the operation's thinking. Instead of trying to improve everything, we focus on the leverage points that matters the most. How to identify your system's true constraint. How to calculate and understand the system's capacity, why speed and availability matters. The difference between coupled and decoupled processes, how strategic buffers protect throughputs. Operation systems are everywhere. They transform inputs into higher value outputs. Products per hour on a factory line, content per week from a creative team. Claims process per day in an insurance office. Patients treated per shift in an hospital. The real question is not where the constraint exists. It always exists. The question is, where can you see it and clearly enough to manage it deliberately. Because once you identify the constraint, you unlock the most powerful lever in operational excellence and take your operations performance to the next level. And that is where the real improvement begins. We usually focus only on KPIs that are showing green, whereas the customer is experiencing the slowdown and the constraint. 4. Bottlenecks vs Constraints — What’s the Difference: Lesson four, bottlenecks versus constraint. And what's the difference? We often use the word bottleneck and constrained as if they mean exactly the same thing. In everyday conversation, they usually do both describe something that restricts or limits the output of the system. The term bottleneck comes from the shape of the bottle. Turn a bottle upside down and try to pour the liquid out as fast as possible. The wide body of the bottle holds plenty of liquid, but the narrow neck controls how fast it can flow out. NC determines the maximum output rate. That image perfectly captures what happens in the system. But if we want to be more precise, there is a subtle difference. The constraint is anything that prevents a system from achieving a higher level of performance relative to its goal. That could be a physical resource such as a machine or lack of material, a policy or a rule that slows the approval process, maybe a budget limit, even a low market demand a bottleneck, on the other hand, is a resource whose capacity is lower than the demand placed upon it. In simple terms, a constraint always exists. A bottleneck exists when the demand exists the capacity. Let's make this practical. Imagine a small bakery that can produce 1,000 pastries per day. But it only receives orders for 600. There is a constraint, market demand. The bakery is sales constrained. However, there is no bottleneck inside the operations because every station has enough capacity to meet the demand. Now, imagine the demand suddenly increases to 1,500 pastries per day. The oven can be 1,200. The packaging station can handle 1,000. The delivery team can transport 900 carefully. Now, the multiple bottlenecks appear. Demand is greater than the capacity at several points. Same bakery, same equipment. The demand level is different. Completely different constraint situation. Here is another example. A hospital has sufficient doctors, available operating rooms, enough nurses to take care, but there are only ten ICU beds. If the demand exceeds those ten beds, the ICU becomes both a bottleneck and a constraint. However, if the patient admission drops below ten per day, the ICU no longer has a bottleneck. The constraint may shift elsewhere. It shows something critical. Constraints are dynamic. They depend on the demand and the capacity. Now consider a more complex case. Suppose two processes in a production line have exactly the same lowest capacity. They both limit the system output. These are joint constraints. If you improve only one of them, the system capacity does not increase. You must improve both. This is why blindly improving busy resources often fails. You must know what is the current demand. What is the capacity for each process? Which resources truly limits the total system throughput? You do not need perfect data. You need enough clarity to identify the constraints and anticipate where it may shift next. So yes, it's useful to understand the technical distinction between a bottleneck and a constraint. But in practice, many professionals use these terms interchangeably. The important part is not the vocabulary. The important part is recognizing what is the limiting performance right now? Because once you identify what is the limiting factor, you know exactly where to focus your efforts, and that is where the real leverage begins. So bottleneck will only arise if your system has a capacity that is slower than the demand. So you may or may not have a bottleneck in the process, but constraints are always present. Just have your eyes open to look for one. 5. Why Constraint Management Matters: Lesson five, why constraint management matters. Constraint management is not just for factories. Yes, many of its early ideas were developed in a manufacturing environment, but the principle applies far beyond production lines. Healthcare system manages patient flow. Hotel manages the room turnover and check in process. IT teams manage the data processing and the server capacity. Agriculture manages harvesting windows and storage limits. Universities manages admission, classrooms and faculty availability. Even at home, families manages time, energy, and resources. Wherever there is a system, transforming inputs into outputs, there is a constraint. So what is the real point of a constraint management? Improving operations sounds good, but it's vague. Improvement must connect to something that is meaningful. In most of the profit businesses, the ultimate goal is simple to make the money sustainable. In the influential book, the goal, Mr. Goldratt introduced the theory of constraint and framed this clearly. The business succeeds by increasing net profits, return on investment, cash flow. And this translates into operational terms. That means increased throughput, reducing unnecessary inventory, reducing operating expenses, throughput is the rate at which the system generates value. It is the speed at which the organization turns efforts into results. For example, a hospital increases throughput by treating more patients safely without increasing the cost. A call center increases the throughput by resolving more customer cases per day without adding headcount. Ecommerce company increases throughput by shipping more orders without expanding the warehouse space. Constraint management focuses on attention to the single point that most limits the throughput. If you increase the efficiency everywhere, except the constraint point, overall output does not increase. But if you improve the constraint point, even slightly, the entire system's performance can improve efficiently and significantly. That is the leverage of constraint management. Instead of spreading efforts, kindly across dozens of improvement initiatives, constraint management directs energy where it has the greatest impact. We will focus primarily on throughput because it is the strongest influence on profitability and performance. When throughput increases, intelligently, profit increases, cash flow improves, and the system becomes healthier. Constraint management is not about working harder. It's about focusing on smarter. And that focus begins with understanding what truly is limiting your system today. A 6. Capacity Constrained or Sales Constrained: Lesson six, capacity constrained, or is it sales? No matter the industry, productive operational systems share the same basic structure. They take in inputs such as material, labor, time, information, or energy. They apply the processes, they generate higher value outputs. That output might be fresh bread from a bakery, a passenger transported safely by a train service, tax filing submitted by an accounting firm. Electricity generated and delivered by the power company. Customer tickets resolved by customer support team, content produced by marketing agency. Every system transforms something into something more valuable. Now, here is the critical question. What is limiting that system today? There are only two broad possibilities. The system is capacity constrained or is it sales constrained? If the business is capacity constrained, it has more demand than it can handle. Customers are willing, orders are waiting, but the system does not have enough capacity to fulfill them. Think of a restaurant with a constant waiting list because there was not enough tables, a popular online course creator who cannot keep up with the enrollment inquiries, a factory with a backlog of orders due to limited machine time, a hospital where appointment slots are fully booked weeks in advance. In these cases, the constraint is operational capacity. The focus must be on increasing the throughput. Now, consider the opposite situation. A company has plenty of capacity, but not enough demand. The bakery can produce thousand loafs per day, but it can only sell 600. The gym has equipment sitting idle for most of the day. The SAS company has survey capacity far beyond the current usage. The consultant has open calendar slots, very few client bookings. In these situations, the constraint is the market. The business is sales constrained, not the operational constraint. Improving production efficiency will not solve the problem in the second case. The attention must shift to marketing, pricing, positioning, and sales effort. Every operational system is in one of these two states and knowing which state you are in is absolutely critical. Many organizations invest heavily in improving internal processes. When the real constraint is lack of demand, other invest in marketing campaigns when the real problem is insufficient delivery capacity. Without clarity, improvement efforts become guesswork. Constraint is not a negative concept. It is an inevitable situation. Every system has at least one limiting factor. Even the most brilliantly managed organizations cannot eliminate constraints entirely. The key is not to remove the constraints altogether. The key is to understand where is the constraint right now and manage it deliberately. Because when you align your actions with the true constraint, performance improves because focused strategic and far more powerful efforts are put in. 7. The Constraint as Your Most Powerful Lever: Lesson seven, the constraints as your most powerful labor. In operations management, understanding constraint is not operational. It is fundamental. Constraints are not something to fear or eliminate blindly. They are powerful leverage points. When managed correctly, they allow you to influence the entire system disproportionately. There is a simple analogy. A chain is only as strong as its weakest link. The constraint is that the weakest link it limits the capacity, it limits throughput. It often limits the sales, and ultimately, it limits the profits. But here's the key shift in mindset. Constraints are not just problems to remove. They are control points to be managed. Even if you eliminate one bottleneck, another constraint will always exist somewhere in the system. By definition, every system has at least one limiting factor. The goal is not to create a constraint free system. That is impossible. The goal is to identify the constraint and deliberately organize the system around it. If a hospital's MRI machine is the constraint, patient scheduling should be built around maximizing its utilization. If a software development team is constrained by testing capacity, work should be released at a pace that the testing can handle it. If a factory paints booth in the slowest operation, upstream production must align to its rhythm. The constraints sets the pace of the entire system. Understanding constraints always affects more than output. It deeply influences the lead time. When constraints are mismanaged, you see delayed deliveries happen. Unpredictable completion dates, long waiting time, frustrated customers and often well intentioned but misguided management actions make these things even worse. For instance, managers may push more work into the system to keep everyone busy. Work in progress increases, inventory piles up. Scheduling becomes chaotic. Urgent jobs are expedited, stress increases, cost increases all while throughput remains unchanged. When inventory increases or when inventory rises without proper control of constraint, the system becomes noisy and unstable. The same principle applies outside the profit driven businesses. A nonprofit organization, the constraint limits how much value can be delivered. A school may be limited by classroom capacity. An aid organization may be limited by the funding cycle. The city traffic flow may be limited by roads capacity at a major intersection. In each case, the constraint determines the rate at which value can be created and delivered. Constraint management teaches a disciplined approach, identify the constraint, optimize it, subordinate other processes to it. Improve it, then repeat as the constraint shifts. These steps are not always intuitive. You must be keen at finding it out. In fact, many a day's management practices directly contradicts the constraint logic. Keeping every resource equally busy feels efficient, but it often reduces the overall performance of the organization. High performing systems does not try to maximize every part. They maximize the flow through the constraint. When you understand this difference, you stop fighting the constraint. You start using it, leveraging it and that shift changes everything from how you manage operations to making it successful. 8. The Airport Throughput vs Throughput Time: Son eight, the airport throughput versus the throughput time. Imagine walking into a busy airport terminal. Announcement echoes overhead. Screen flashes, departure times, passengers move with urgency towards check in counters, security lanes and boarding gates. Without realizing it, you have just entered one of the most sophisticated operational systems in the world. An airport is a people processing system. Its core flow includes check in or baggage drop, security screening, boarding, and takeoff. From the passengers point of view, the experience can feel long, bureaucratic and stressful. Two to 3 hours from entry to takeoff is common at an international travel. But for an operation's perspective, something very different is happening. Most major airports are capacity constrained. The constraint is almost always the runway. Runways are extraordinarily expensive. Space is limited. Regulations strictly governs the aircraft separation, safety distance, and movements per hour. Expanding the runway capacity is rarely simple and often impossible. The runway determines how many planes can depart per hour. It determines the throughput. To maximize the system output, airport managers must keep that runway fully utilized at all the times. That queue of aircraft waiting to take off is not poorly planning. It is deliberate inventory buffer. It protects the runway from upstream variation. If the check in slows down or the boarding is delayed, the runway must still remain busy. Idle runway time means the loss throughput. Now, here's the important distinction. Passengers care about throughput time, how long it takes to move through the system. The airport cares about the throughput, how many passengers departed per hour. These are not the same. Imagine two pipes with identical diameters, but different length. Both can push out the same number of units per minute. One pipe is long, the other is short. Throughput is identical. Throughput time is different over here. From a passenger's perspective, the throughput time is frustrating because it's long. From the airport's perspective, throughput time is not the primary concern. In fact, airports are unusual systems. Approximately 40% of the airport revenue comes from non aeronautical activities, such as retail, dining, property rentals. The longer the passenger remains inside the terminal, the more likely they are to spend. In this specific case, the longer throughput time can actually increase the revenue for the airport. Contrast this with a factory. In manufacturing, long throughput time increases work in progress. Ties up the capital increases the storage cost, reduces responsiveness. Factories do not benefit from long internal flow time. This example highlights a powerful lesson. You must understand the system's goal before deciding what to optimize. For an airport, maximizing runway utilization is definitely important. For a factory, reducing WIP or work in progress, shortening the lead time is important. For a hospital, maximizing safe patient throughput. From a retail store, balance customer experience with transaction speed. Sunstrain management is always contextual. Next time, you wake up early for a flight, move through the check in, security, duty free shops, and boarding gates. Remember, you are flowing through a carefully orchestrated operational system and somewhere behind the scenes, the most critical role is being played by the air traffic controller. The guardian of the runway constraint, ensuring maximum throughput while keeping everyone safe. That single constraint quietly determines the performance of the entire airport. I guess you learned a new perspective in this lesson. I will see you in the next lesson. 9. The Theory of Constraints and the Five Focusing Steps: Lesson nine, the Theory of Constraints and the five focus steps. Much of the modern bottlenecks and constraint management is built around the theory of constraint, often called as TOC. TOC is a powerful system. Improvement methodology is based on one central idea. The system's output is limited by its constraint. If you want to improve the output or the performance of the system, you must identify that constraint and manage it deliberately. Oh the Theory of Constraints developed by IlahuGolut and popularized in his influential book, The goal. Although it began in manufacturing, Theory of constraint applies far beyond factory floors. It has been successfully used in job, batch, mass, and continuous production, supply chain and logistic, finance and accounting, marketing and sales, healthcare and service operations. We will focus on the application of this process. What TOC optimizes. In operational term, theory of constraint encourages us to increase the throughput, reduce the inventory, reduce operating expenses, but the primary focus is throughput. Importantly, TOC defines throughput very specifically. Throughput is money generated from actual sales minus variable cost. Producing goods that sit in storage does not increase the throughput. Filling a warehouse is not a success metric. Only what is sold is counted. This mindset shift attention aways from local efficiencies towards system wide flow. Theory of constraint teaches us to balance the flow, not the capacity. The five focusing steps, the most famous TOC tool is the five focusing steps. Once the system is mat, the goal is cleared. Improvement follows the sequence. Step one, identify the constraint. Which process currently limits the overall throughput? Where is the weakest link? Step two, exploit the constraint. Make sure that the constraint is fully utilized, reduce the downtime, eliminate unnecessary interruptions, simplify the procedure around it. Keep it running effectively. Step three, subordinate everything else to the constraint. All other processes must align with the constrained space. Do not overproduce the upstream. Ensure support functions prioritize the constraint needs. Step four, elevate the constraint. Only after fully exploiting and subordinating should be considered about adding capacity. This might mean buying another machine, hiring additional staff or redesigning the process. Step five, repeat the cycle. Once the constraint is broken, it will shift elsewhere. Identify the new constraint and begin again. The most importantly, do not let inertia become the new constraint. The brilliance of theory of constraint lies in its discipline. Instead of trying to improve everything at once, it channels the effort into single point that matters the most. It replaces scattered improvement initiatives with focused leverage. And when applied consistently, it transforms how the system performs. 10. Start with a Process Map: Lesson ten. Start with a process map. When improving any operations, the first step is simple. Map the process. Before jumping into solutions, draw a clear picture of how work flows through the system. Keep it simple at first. Break the operations into three or ten major steps. You can always add details later once you know where to focus. Let's take a relatable example. Imagine a fast food sandwich bar making toast foot long sandwich. From the shop's perspective, the process might look like, take the customer's order, add the meat and the cheese and the vegetables, toast the sandwich, add some salads, and the toppings, wrap the sandwich, take the payment, add the till. Each step takes different amount of time. And because the time determines how many units can be completed per minute, each step has its own capacity. Simplify, assume one employee per step. No shared resources, no breakdowns or delays. Perfect availability. In this ideal scenario, process time directly determines the capacity. Now imagine the toasting step. It takes 40 seconds while all the other steps take less time. It means toasting can only produce 1.5 sandwiches per minute. Even if every other station works faster, the entire system cannot exceed that rate. Toasting is a constraint. Three toast in 2 minutes. Now, suppose the owner buys a second toaster. At first glance, it seems like the system's capacity will be doubled, but it will not. Yes, toasting capacity has increased. But once the constraint is relieved from this station, the next slowest process becomes the new constraint. Perhaps salad assembly, perhaps maybe the payment at the til the constraint now shifts. This is why process mapping is very important. To identify the true constraint, you must map the flow. Estimate or measure the capacity at each and every step. You can do a time and motion study. Compare demand to capacity. Only then can you clearly see what limits the system? Without a process map, improvement efforts are guesses. With one, constraints identification becomes logical and data driven. And once the constraint is visible, meaningful improvements can begin. I 11. When Systems Get More Complex: Lesson 11, Let's move beyond a simple linear process. Imagine a system with four processes, and theoretically, capacity is seven units per R. That means somewhere in that system, the constraint limits the output to seven. Now, suppose we buy another machine for process two and double its capacity 7-14. It feels like a major upgrade, but the system does not double instead, the constraint shifts. Process two is no longer the bottleneck. Now, the constraint moves to the two joint processes that begin and end at the system. To increase the capacity further, both of those processes must be improved together. This illustrate a critical principle. Improving one part of a system does not guarantee a system wide improvements. Now, let's add more realism. Imagine the system has two separate input streams feeding into a common process. Each input stream has its own capacity and its own internal constraints. If the top input stream is limited to eight units per, then the overall system cannot exceed eight regardless of what happens the downstream. If that stream is increased to 12 units per hour, the constraint may shift to a merging point where both the streams meet, and perhaps now the limit to ten units per hour. In multi input systems, the slowest feeder can limit everything to operate at a target capacity. All the input streams much support that level. This is where simplification becomes powerful. Instead of analyzing every microstrips, you can collapse each input stream into a single effective capacity number. That allows you to see the systems more clearly and identify where the real constraint lies. But it gets even more interesting. Suppose the system produces three different product variants. All variants pass through the early stages of the system, but then split into three separate finishing processes. Each final process may have a small individual capacity. However, their combined output might appear sufficient. At the first glance, you might think the constraint lies in those final processes. But if all the variants share the same upstream process that is limited to ten units per hour, then that first shared process is the true constraint. Even if the downstream branches can handle 15 units combined, it will never receive more than ten. The shared upstream resources governs the total system output. This is why simply looking for the smallest number is not enough. You must understand shared processes. Parallel flows merging points, product mix and variance. Real world systems are rarely linear. They involve multiple inflows, branching paths, shared resources, and variant outputs. The key is to map the system. Assign capacity estimates to each process, simplify branches into effective capacities. Identify the constraint at the correct level of resolution. Then zoom back to where the action is required, start broad, simplify and identify the constraint, then dive deeper only where necessary. Constraint management is not about drawing complicated diagrams. It is about creating clarity, and that clarity allows you to focus your efforts exactly where it will change the system's performance the most. 12. Coupled vs Decoupled Processes: Lesson 12, coupled versus decoupled processes. Not all the processes in a system behaves the same way. Some processes are coupled, the others are decoupled. Understanding the difference is essential for effective bottleneck management. What are coupled processes? Coupled processes are interdependable processes. They rely on each other to operate. If one stops, the other must stop. Imagine two machines in a production line connected to a conveyor belt. They must run together. If the upstream machine stops, the downstream machine must have nothing to process. If the downstream machine stops, the upstream machine has nowhere to send the output. They are tightly linked. That is called as coupling. Sometimes coupling happens for physical reasons. Sometimes it happens for quality, safety and procedural reasons. Take a sandwich shop. Once the sandwich is toasted, it must immediately move to the filling station. The business cannot toast a large batch in the night before and store it. Freshness requirement effectively coupled the toasting and the filling process. If the toaster breaks, filling stops. If the filling stops, the toasting must stop. Now consider a hospital, an operating theater and the recovery ward are coupled processes. If the operating theater is unavailable, the recovery. The recovery ward has no patients. If the recovery ward is full, surgeries must stop because there is no where the patient can go. These human centered processes are just tightly coupled as machines. Coupling can also arise from the rules or the policies. For example, a dangerous industrial process may require supervision before proceeding. That supervisory approval couples two activities that might otherwise operate independently. Couple process affects the throughput throughput time, lead time, flexibility, Dependency is not always obvious. They may exist because of legacy procedures that no one has questioned in years. Coupling is not inherently good or bad. Advantages of coupling, lower inventory, simpler coordination, faster immediate flow, disadvantages. Higher vulnerability to downtime, reduce flexibility, greater risk of stopping the constraint. If a constrained process is tightly coupled to an unreliable upstream process, every disruption upstream reduces the system throughput. What about decoupling? Decoupling separate the process. They can operate independently. This is typically done by allowing the inventory or work in progress between them. That inventory when deliberately planned and managed is called as a buffer. A buffer is not a random pile of inventory. It is a strategic protection. For example, in a manufacturing, a small buffer before a constraint, ensure it never runs out of work. In healthcare, scheduled patient flows ensures operating theater is continuously utilized. Buffer absorbs variability. However, unmanaged inventory is not a buffer. It's a waste. The key difference is intention and control. Whether to choose a couple or a decoupled process depends upon whether the process is a constraint due to safety requirements, quality requirements, cost considerations, competitive priorities. In high variability environment, decoupling may protect the throughput in a stable environment, coupling may reduce the cost and the inventory. The real question is not is coupling good or bad. The real question is, does this dependency protect or threaten the constraint? Because in constraint management, protecting the constraint is always the priority. Let us now understand customer service at a bank as a coupled process, imagine you go to a bank to update your address. Step one, the front desk officer collects your form. Step two, the verification officer checks your ID documents and approves it. Now, suppose the front desk officer cannot submit your request unless the verification officer is immediately available to validate and approve it. If the verification officer is busy or absent, the front desk must stop processing the address update on that day. Both the steps must happen together at the same time. If one stops, the other stops. This is coupling. It's like two kids doing a three legged race. If one trips, both fall. Now imagine the bank changes the system. The friend desk collects your form, uploads it in a digital queue. The verification officer checks the request later in batches. Even if the verifier is busy for an hour, or the friend desk can keep accepting the forms. There is a small waiting list between them. That waiting list is a buffer. Now, the two processes can work independently. This is decoupling in action. It's like putting homework in a basket for the teachers to grade and she can do it later. Students can keep submitting work even if the teacher isn't grading at that very moment. Great. Now, let's imagine from a telecom company example, you call your telecom provider because your Internet is not working. Step one, the call center agent listens to your issue. Step two, the technical team resets your connection in the Ben. If the agent must wait on the line while the technical team fixes it, immediately, both processes are tightly linked. If the technical team is unavailable, the agent cannot close the call. If the agent disconnects the technical team does not start, they depend on each other at the same time. This is an example of coupling. It's like two people carrying a table. If the one lead goes, the table will drop. The goal is not to remove all the coupling. The goal is to protect most important step in the system, the constraint. Here's how you do it. Stepan ask an important question. Which system limits how many customers we can serve. In the example of the bank, if the verification can only approve 20 requests per hour, but the front desk can take 50, then verification is the constraint. In the scenario of the telecom company, if the technician from the back end can only fix 30 issues per hour, but the call center receives 60 calls, then the technical team is the constraint. Once you know this, everything becomes easy and clear. Step two, protect the constraint. Now we decide whether to couple or decouple it. The rule number one is never let the constraint sit idle. If the technical team is the bottleneck, they must always have work ready. So you create a ticket queue, incense a buffer. Make sure that the requests are clearly written, remove unnecessary approvals. You decouple the upstream process so that the constraint never waits. Step three, remove harmful coupling. If the coupling causes the constraint to stop or break it or delay it, you must fix it. If I go back to the friend desk of the bank, the friend desk must wait for the verifier to approve each request immediately. The solution would be allow the request to be logged first. Verification happens later. Create a digital queue. Now, the friend desk keeps working. Verification always stays busy. The throughput increases. If I talk about the other problem that we just saw, the agent must stay on the call until the back end team resolves the issue in the telecom company. We can create a solution by creating a ticketing system. The back end resolves issue separately, customer gets SMS update. Agent moves to the next call. Now, the backend team, which was a constraint now works continuously. Agents don't block. Customer gets faster response overall. Step four, sometimes you keep coupling on purpose. If the coupling protects the quality or the safety, you keep it. In the bank loan scenario, approval might require fraud check before releasing the loan. You don't decouple that blindly. Does this coupling protect the quality? Does it just slow down things for no reason? Is the questions you ask. Step five, improve the constraint itself. After organizing the flow, improve the constraint by using better tools, providing training and automation, if possible. Remove paperwork because that's where the rework will increase, and by removing paperwork, you reduce the rework. Never buy more people or more systems first. First, you must organize, then you improve. Then we think about investment because ROI is important. The simple formula is, find the slowest step first, keep it busy all the time, remove dependency that stops it. Add a buffer before it before the constraint, improve it gradually. In a single sentence, you don't solve coupling or decoupling. You design a system so that the most important step never stops. That's constrained thinking. That's called a exploiting the constraint. Now imagine the company changes the system. The agent logs your complaint into a ticketing system. The technical team works on the tickets in an order. The agent can take the next customer call immediately. The ticket queue between them is the buffer. The two teams no longer need to work at the exact same moment. That is decoupling in action. It's like placing the food order at the restaurant counter. The cashier keeps taking orders even if the kitchen is still cooking the previous one. Coupled processes must work together at the same time. Decoupled processes can work separately because of something called a buffer that connects them. The smart question is, operation is always busy and does this connection protect the most important step in the system? Or does it make everything stop too easily? 13. Spotting the Constraint in the Real World: Lesson 13, spotting the constraint in the real world. Mapping processes and calculating capacity is essential if you want precision. But sometimes, before the spreadsheets and the analysis begins, you can learn a lot just by observing what is happening. When operations are under pressure and deliveries are falling behind, it's common to hear everything is a bottleneck. Technically, if the demand exceeds the capacity for the process, everything could appear as a constraint. But in reality, that situation is rare. There is almost always one primary constraint and perhaps two closely linked ones. The challenge is to spot the early queues. The most obvious clue is that a growing queue. If the system where the processes are not tightly coupled, the constraint often reveals itself very clearly. As a consultant, you should look for large and growing queues of work waiting in front of a process. That process already working at its maximum effort, the upstream process is finishing work faster than it can be handled. In a call center, you may see calls tacking up in the queue, while the agents in one specific team are constantly busy. In a warehouse, you might see packet accumulating before a packing station. That is always operating at a full speed. In an HR department, offer letters often pile up waiting one final approval signature. The growing queue in front of a fully utilized process is a strong indication of a constraint. It's not a foolproof way, but it's highly suggestive. When everything looks blocked, what should you do? Now, consider a tightly coupled system. In a fully coupled production line, inventory cannot easily accumulate. At a single visible point, if one process slows, the entire line slows down. Think of a city grid locked with the traffic. Cars are backed up everywhere. Every road looks congested. Even from above, it can be difficult to immediately identify the root cause. But if you had been watching earlier, you might have noticed where the traffic jam first began to build. That initial point of accumulation often signals the constraint. Timing matters, observation matters, and hence you should build this skill of observing the process from a distance. Look for the highest utilization. In a tightly linked system, another powerful indicator is utilization. The constraint is often the resource that is busiest has the least idle time. This is constantly under pressure. Experience the most expedity. In a hospital, the MRI machine may run continuously with no downtime. In a software team one senior reviewer may always be busy or overloaded with approvals. In a data center, one server cluster may run at consistently high CPU utilization. This is your hotspot. Think of using a thermal imaging camera for a circuit board. The hottest component is often the one that is working the hardest and potentially limiting the performance. Talk to the people who are doing the work. They will be able to guide you. Data is powerful, but it is not enough. Operators, supervisors, schedulers, maintenance team, and frontline staff often have invaluable insights. They know where work regularly backs up. Which process always feels rushed, where urgent jobs get expedited, where breakdown causes the most disruption. It is rare for any one person to see the entire system. But collectively, insight is powerful. You must combine observation with practical experience and basic utilization of data, structured constraint thinking, and you will narrow your focus quickly. Before dividing into complex calculation, walk the floor. Gemba is great. Ask questions, watch the flow. Constraint identification is both analytical and observational. When you blend discipline analysis with hands on inside, you move from guessing to understanding, and that is where the real improvement becomes possible. 14. Inventory Management Through the Lens of Constraints: Lesson 14, inventory management through the lens of constraints. Inventory management is a core of operations management, but it is not just about stocking elves, managing warehouse, or ensuring deliveries arrive on time. Inventory plays a critical internal role in enabling maximum throughput without allowing costs to spiral out of control. There are different types of inventory. In most system, inventory falls into few main categories. Raw material that is entering into the system. Work in progress or WIP once processing has begun. Finished goods which are ready for customers to use. Plus spare parts and consumables that support the operations. When it comes to constraint and bottleneck management, work in progress is the most critical one and often the most neglected one. It is also the complex one. Work in progress is essential for keeping the constraint running, but it is also the easiest type of inventory to let grow out of control. The danger of keeping everyone busy. In many operations, processes are not tightly coupled. That means they can operate independently provided they have enough work available. At first glance, this seems positive. In every process always has something to work on. Productivity should increase. But when used carelessly, this thing leads to an increase in work in progress, long and unpredictable lead times, sgested workplaces, impossible scheduling, high holding, and financing cost. Inventory begins to spread everywhere in the system, and ironically, throughput does not increase. What is the constraint based approach to inventory management? When managing inventory through a constrained lens, the principle is clear. Ensure that the constraint always has input inventory it needs while minimizing the overall system inventory. This usually requires deliberate inventory buffers placed in strategic location. A buffer is not a random pile up. It's a planned sized and managed inventory designed to protect the constraint from variability. The goal is to protect the throughput, not to maximize the stock. The lean principle of pull and carbon. The lean philosophy is built on the principles from the Toyota production system. It emphasizes on flow, waste reduction, inventory reduction. These ideas are highly compatible with the constrained thinking. When applied with focus, one of the most important lean concept is pull. Pull means that the downstream demand triggers the upstream production. Work is pulled through the system rather than pushed in anticipation. This reduces overproduction, one of the biggest driver of excessive WIP. A well known pull mechanism is Kanban. Kanban uses visual signals, physical or digital cards that travels with the work item. When a job is completed, the CR Ban signal returns upstream to authorize new work. This naturally limits the WIP. It is also important to note that the modern agile Kanban boats share a similar concept, but they are not identical to the traditional production Carvan system. There are many inventory control approaches, including just in time, rum, buffer rope, C, WIP, tube in systems, FIFO lanes, and various carbon variations. Mostly rely heavily on visual management to make inventory visible and controllable. The core principle is that in a capacity constrained system where maximizing output is the goal, Inventory management must play a critical role, and it should support the constraint in maximizing the throughput. That means that the constraint should never wait for the material or information. BFR should protect it from variability. Inventory elsewhere in the system should be tightly controlled. The objective is a balance. Too little inventory and the constrained staves and too much of inventory, then we create a bottleneck and the system clocks. Constraint based inventory management prevent both the extremes. When done correctly, it protects the throughput while avoiding the costly consequence of uncontrolled WIP. That balance is where operational excellence truly lives. Now, imagine this problem in a bank. The customer applies for a loan. The steps it follows is that the front desk collects the documents. The credit officer checks the documents. The manager approves the loan. If the credit officer can only check 20 files per day, but the front desk collects 60 applications, something happens. There is an inventory pile up. If the front desk keeps sending all the 60 files immediately, the files pile up on the credit officer's desk. The officer feels stressed, approvals get delayed. Everything is tightly linked. If the credit officer slows down, the whole system slows down. That's coupling. Too many files or WIP is equal to the system getting clogged. Let's move to a smart buffer solution. Instead of sending unlimited files, the bank allows only 25 applications into the check in queue. New applications wait in a controlled digital queue. The credit officer always has work, but not too much of work. Now, the credit officer or the constraint never waits. Files don't pile up everywhere. The processing becomes smoother. That control queue is a buffer. It's a planned protection for the constraint. Imagine the customer calls the telecom support. The call center agent locks the complaint. The technical back in team fixes the issue. If the agent must stay on the call until the backend fixes the issue, the agent cannot take new calls. The backend team becomes the overload. Customer waits longer. So if the back end team fixes only 30 issues per hour, but the agent receives 60 complaints, the ticket pile up. There is too much of work in progress in the system and everything slows down. So what could be a smart solution? Here, we are going to talk about ticket buffer. The agent creates support tickets. Only a limited number of tickets are allowed in the ready to fix cube. The backend team always has tickets to work. Customer receives SMS update when the issue gets resolved. Now, the backend team, which is a constraint, never waits for work to arrive. At the same time, they are not overwhelmed because tickets don't pile up endlessly. Agent keeps taking new calls that limits the ticket queue entry. The buffer is protecting the constraint. The simple lesson that we learned is coupling is equal to everything depends upon each other tightly. Too much work can clog the system. The solution is adding a control buffer so that the slowest step or the constraint never stops. It does not have too little inventory for the constraint to stop. At the same time, there is not too much of inventory for the system to feel clogged. Smart inventory is equal to just enough to protect the constraint. That's constraint based inventory management. 15. Lean, Six Sigma, and Constraint Thinking: Lesson 15, Lean Six Sigma and constrained thinking. Lean and Six Sigma are two of the most widely used operational improvement approach in the world. Both are powerful. Both are respected, and both are highly compatible with bottlenck management and constraint management and the theory of constraint. But there is an important distinction between lean and Six Sigma. Both provide excellent tools on how to improve a process. Constraint management helps you decide what to improve. That difference is critical. Not all processes matter equally. One of the central lessons of constrained thinking is that not all processes contribute equally to the system performance. Improving a non constrained may make the team feel productive. It may reduce the local waste. It may increase local efficiency, but it will not necessarily increase overall throughput. Choosing where to focus is more important than choosing how to improve. So where does Six Sigma fit in this situation? Because Six Sigma is designed for rigorously identifying and reducing variation and defect in the process, it emphasizes on a data driven decision making, statistical analysis, root cause identification, process control. In the capacity constrained systems, applying Six Sigma two directly to the constraint can be extremely powerful. If a packaging line is constrained, frequent stop due to quality defects, Six Sigma can reduce those defects and directly increase the throughput. But if the Six Sigma project are applied to the process with excess capacity, the impact on the overall output may be minimal. The tools are strong, the direction of focus determines its impact. Lean is a broader operational philosophy. It is based on eliminating waste and improving flow. It offered tools such as value stream mapping, Kison, standardized work, visual management, continuous improvement practices. Lean emphasizes on flow aligns closely with constrained thinking. However, Lean effort sometimes fail into the trap of trying to eliminate waste elsewhere without prioritizing. Improving non critical process may generate energy and engagement, but without focusing on the constraint, the bottom line results can be underwhelming. Lean strongly advocates reducing the inventory, and in many cases, that is beneficial. But constraint management teaches that strategic inventory buffers can protect the throughput. Blindly eliminating all the inventories can unintentionally expose the constraints to the upstream variability. Reduce the system performance. The key is intention. Uncontrolled inventory is waste. Deliberately managed buffer that can protect the constraints are strategic tools. You do not have to choose the sites. Improvement methodologies are not competing religions. You do not need to pick a ham. Lean, Six Sigma, and Theory of Constraints are largely compatible with each other. The most effective approach is use the constrained thinking to identify the critical leverage point. Use Lean to improve the flow. Use Six Sigma to reduce the variation and defect at it. When applied together with a clear understanding of the constraint, these methods reinforce one another. Without focus, improvement is shattered. With constraint based focus, improvement becomes powerful and the tools matter. But where you apply them matters far more. Let's make this very practical and real world. I'm going to take up examples from FedEx, Amazon Ecommerce, and a steel sheet manufacturing company, where I will focus on explaining what can Lean and Six Sigma improve? What constraint actually is, and why focus matters more than tools. So FedEx, as you know, is known for package sorting. The situation is at a large FedEx sorting hub, trucks arrive with packages. Packages are scanned. They are sorted by the conveyor system. This is loaded onto the outbound trucks or planes based on the QR code. Let's say the aircraft departure schedule is the constraint. Only ten flights can leave tonight. That limits the total number of shipments. The lean team improves office paperwork process, breakroom organization and shrinks the scanning time from 3 seconds to 2 seconds. Everything looks better, but plane still departs at the fixed time. Throughput doesn't increase. Local efficiencies improve. System output did not improve. Is it worth your efforts? Can you call this project a success? Now comes the constraint based thinking where we focus what we want to solve. The aircraft loading time before the departure. Now apply tools correctly, improve the flow on the loading dock, remove the wasted motion, standardize the loading sequence. I can use Six Sigma over here to reduce the loading error that causes rework. We might also focus on reducing the mis loaded packages. The result would be that the plane leaves full and on time. Me packages are delivered throughput increases. Same tools, different focus, and very different impact. Now, think about an Amazon ecommerce fulfillment center. The customer order flows when he places the order. The item is picked, it is packed, shipped. Suppose that the packing stations are the constraint. They can pack 500 orders per hour. The pickers can pick 700 orders per hour. So you know very clearly where is the constraint. The lead team reduces the walking distance for the pickers. It's a great improvement. The picker now picks 800/hour, but the packing still is only at 500. What happens? The WIP piles up, inventory grows, and the stress increases. The throughput still stays at 500. If I go with a constraint based thinking, I'm going to now apply the tools differently. I will identify the constraint, that is the packing department. I'm going to apply the Six Sigma concept to the packing department. I will focus on reducing the packing errors, reduce the box size mistake, and reduce the repacking. I might also use lean concepts like improving the layout, pre stage boxes using visual management. The result is that the packing now increases 5-650/hour. Now, the system throughput has increased. Improving the picker didn't help, but improving the constraint definitely helped. Now think about a steel sheet production flow. The furnace melts the raw material. The rolling mill shapes the sheet. Cooling, followed by cutting and inspection, that is the flow of the process. Now suppose the rolling mill is the constraint. It is extremely expensive and a slow. Everything must pass through the process. Now, if I do a Six Sigma project on inspection, the inspection team runs a Six Sigma project. They reduce the defect by 30%, creates a local improvement, but the rolling mill capacity is unchanged. Total tons per day remain the same. There is no increase in the output. Applying Six Sigma on a rolling mill instead would be the right thing to do. It reduces the downtime, improves the cycle time, and reduces the setup time. It also reduces the variation in thickness. Lean can focus on ensuring that the material is always ready before the rolling. It will remove the waiting time and optimize the single minute exchange of dy. The rolling met now produces more tons per day. The entire plant throughput increases because everything flows through it. Wrong focus will give you wrong result, whereas a right focus will give you the right result. Lean and Six Sigma can answer, how do we improve this process? Constrained thinking answers the question, which process should we improve first? Without constrained thinking, improvement is scattered. With help of constrained thinking, improvement multiplies and we can see the result.