Process Analysis, Bottlenecks & Throughput

Almost everything a business does is a process — a sequence of steps that turns inputs into outputs. Making a bicycle, approving a loan, serving a coffee, onboarding an employee: all processes. The first skill of operations is learning to see processes clearly, measure them, and find the one step that limits the whole thing — the bottleneck. This lesson gives you that diagnostic eye, plus Little's Law, a deceptively simple equation that governs every process on Earth. Forge Bicycles' production line is our workbench.

Part 1 — Everything is a process

A process transforms inputs into outputs through a series of steps: inputs → transformation → outputs. The foundational operations skill is process mapping — drawing the flow of steps (a flowchart) so you can actually see where work goes, where it waits, and where it gets stuck. You can't improve what you can't see, and most operational problems become obvious the moment someone maps the process honestly.

Forge's bike assembly has four sequential stations:

Frame prep → Wheel building → Assembly → Quality check + packaging

Each station processes bikes at its own rate. Simple enough — but how fast can the whole line produce bikes? The answer is not obvious, and getting it right is the heart of operations.

Part 2 — The metrics that describe a process

Four numbers describe any process; learn them and you can diagnose almost anything:

• Throughput (rate) — units completed per unit of time (bikes per hour). The output of the process.
• Flow time (lead time / cycle time) — how long one unit takes to travel through the whole process, start to finish.
• Capacity — the maximum throughput a step (or process) can sustain.
• Work-in-Process (WIP) — the number of units inside the process at any moment (started but not finished).

Notice that throughput and flow time are different questions: throughput is "how many come out per hour?"; flow time is "how long does one take?" A process can have high throughput but long flow time (a busy factory with lots of WIP) or low throughput but short flow time. Keeping them distinct is essential — and Little's Law (Part 4) ties them together.

Part 3 — The bottleneck rules everything

Here's the single most important idea in process analysis. The throughput of an entire process is limited by its slowest step — the bottleneck. Like a chain that's only as strong as its weakest link, or a highway whose flow is set by its narrowest lane, a process can only move as fast as its tightest constraint.

Look at Forge's four stations and their individual capacities:

Even though three stations can handle 20–30 bikes/hour, the line's throughput is capped at 15 bikes/hour — the rate of the wheel-building bottleneck. The faster stations simply sit partly idle, waiting; inventory piles up in front of the bottleneck and starves the stations after it. Forge cannot make more than 15 bikes/hour no matter how hard the other stations work.

The counterintuitive consequence: speeding up a non-bottleneck station does nothing for total output — it just creates more idle time or more pile-up. To produce more bikes, Forge must improve the bottleneck (wheel building) and only the bottleneck. Effort spent anywhere else is wasted.

Part 4 — The Theory of Constraints

This insight scales into a whole management philosophy — the Theory of Constraints (Eliyahu Goldratt): every system is limited by a small number of constraints, so to improve the system, focus relentlessly on the constraint. The recipe:

1. Identify the constraint (the bottleneck — Forge's wheel building).
2. Exploit it — make sure the bottleneck is never idle or wasted (it's the most precious resource; don't let it wait for parts or work on defective units).
3. Subordinate everything else to it — run the other stations at the bottleneck's pace, not flat out (running them faster just builds useless WIP).
4. Elevate the constraint — add capacity to it (another wheel-building station, automation, a second shift).
5. Repeat — once you fix one bottleneck, a different step becomes the new constraint, and you start again.

That last point matters: if Forge raises wheel building to 22 bikes/hour, the bottleneck moves to assembly (20/hour), which now caps the line. Improvement is a never-ending hunt for the current constraint. This is why great operations is a culture of continuous improvement (Lesson 2), not a one-time fix.

Part 5 — Little's Law: the universal equation of flow

One elegant relationship connects the three flow metrics, and it holds for any stable process — a factory, a hospital, a checkout line, a software backlog:

Little's Law

Work-in-Process (WIP) = Throughput × Flow Time (often written L = λ × W — inventory = arrival/throughput rate × time in system)

It's almost magically general. Three quick uses for Forge, running at 15 bikes/hour:

• Find flow time: if there are 45 bikes in process (WIP) at a throughput of 15/hour, then flow time = WIP ÷ throughput = 45 ÷ 15 = 3 hours. A bike takes 3 hours to traverse the line.
• Reduce WIP → reduce flow time: at the same 15/hour throughput, cutting WIP from 45 to 30 bikes drops flow time to 2 hours. Less inventory in the system means faster delivery — the same product comes out quicker.
• The cash connection: WIP is inventory, and inventory is cash tied up (Modules 1–2). Lowering WIP shortens flow time and frees cash and speeds delivery — a triple win, which is exactly why lean operations obsess over reducing WIP (Lesson 2).

Little's Law is the quiet workhorse of operations: give it any two of {WIP, throughput, flow time} and it gives you the third. Internalize it and you can reason about the flow of any system — including the queue you're standing in.

Key takeaways

• Almost everything is a process (inputs → transformation → outputs); process mapping makes problems visible.
• Describe a process with throughput (units/time), flow time (time for one unit), capacity, and WIP (units inside the process).
• The bottleneck (slowest step) caps the whole process's throughput — so improving a non-bottleneck does nothing; improve the bottleneck only.
• The Theory of Constraints — identify, exploit, subordinate to, and elevate the constraint, then repeat (the bottleneck moves once fixed).
• Little's Law: WIP = Throughput × Flow Time holds for any process (Forge: 45 bikes ÷ 15/hr = 3-hour flow time); reducing WIP cuts flow time and frees cash.

Knowledge check

1. Forge's stations run at 25, 15, 20, and 30 bikes/hour. What is the line's throughput, and which station is the bottleneck?
2. Forge speeds up its QC station from 30 to 40 bikes/hour. What happens to total output, and why?
3. State the five steps of the Theory of Constraints.
4. Using Little's Law, if Forge has 60 bikes in process and a throughput of 15 bikes/hour, what is the flow time?
5. Why does reducing work-in-process (WIP) benefit a company in three ways at once?

Finding the bottleneck tells you where to improve. The next question is how — by systematically removing waste and defects. Next: lean, Six Sigma, and the quality revolution that built Toyota.

→ Lesson 2: Lean, Six Sigma & Quality