What Is Throughput and Why You Should Care

Nov 6, 2018 | Manufacturing, Theory of Constraints

Throughput is the #1 metric for assessing the quality of your production line. Not just an important metric, but the most important metric.

In business-bottom-line terms, throughput is the difference between:

  • Meeting your production goals and missing your targets
  • Having a competitive advantage and falling behind
  • Keeping your customers and losing them to someone who can produce more / better / faster

If you’re a plant manager, a process improvement engineer, or someone else whose compensation is tied to production, throughput is the difference between a nice fat quarterly bonus and having to dial back those vacation plans.

Given all this, you might be surprised to know that most production lines, especially in the food and beverage manufacturing industry, are seriously underperforming on this metric. As in, their throughput is a good 20-30% lower than it could be. And, that’s just the average — we’ve seen production lines with room for up to 60% improvement in their throughput. Can you imagine what a 60% improvement would look like?

The reason production lines underperform is because throughput too often takes a backseat to efficiency. Now, don’t get me wrong — efficiency is important. But if your goal is to maximize the amount of product coming off of your line every shift, efficiency alone won’t get you there. You could have every piece of equipment running at peak efficiency and still have suboptimal throughput.

Let’s dive into why that is and, most importantly, what you can do about it.

What is throughput?

Simply put, throughput is a way to measure the effectiveness of your entire production line. In technical terms, it’s the rate of production, i.e., how much you can produce over a certain period of time.

Throughput = Units Produced / Time

If a wine bottling line produces 6,000 bottles of wine an hour, then its throughput is 100 bottles per minute.

There are two types of throughput we can calculate:

  • Throughput of individual machines
  • Throughput of the entire line

The reason most production lines underperform is that companies only consider the throughput of individual machines. But it’s the throughput of the entire line that determines whether or not you meet your production goals.

Throughput of individual machines

On a machine-by-machine basis, throughput is highly related to efficiency, which is equivalent to uptime percentage.

In a perfect world, every machine along a production line would run nonstop, which means the uptime percentage would be 100%. But, in the real world, machines fail all of the time. Going back to our example of the wine bottling line, maybe the labeler runs out of labels, or the foil in the capsuler gets jammed, or someone trips and accidentally shuts off the decaser.

When these things happen, it takes time to get the machines back up and running again. We can use the mean time between failures (MTBF) and the mean time to repair (MTR) to calculate a machine’s probability of run (POR).


In English:

Uptime % = Uptime / Total Time

Move the decimal a couple of places, and you have the machine’s efficiency.

It may sound complicated, but, once we put it into an example, you’ll see that it’s actually quite simple. Suppose the decaser on our wine bottling line can decase 240 bottles per minute (bpm). It takes 1.5 minutes to repair and the mean time between failures is 30 minutes.

First, we calculate the probability of run:

POR = 30 / (1.5+30)
POR = 30 / 31.5
POR = 0.9523

For simplicity, let’s call it 0.95. Then, we move the decimal point two places:

Efficiency = 95%

Now, we calculate throughput by multiplying the POR by the decaser’s maximum rate:

0.95 x 240 bpm = 228 bpm

I know what you’re thinking — 95% efficiency is very good. And you’re right! No machine runs at 100% efficiency, so we’re likely getting the maximum throughput out of this machine already. Case closed!

Not so fast.

This is where most companies stop. They analyze their individual machines, see high efficiency numbers, and call it good.

And that’s exactly why their lines underperform.

To understand why, we need to perform a line analysis

Throughput of the entire line

So, we’ve established that the decaser is working as hard as it can. That’s great! But the decaser isn’t the only machine on the line. There’s also a filler, a capsuler, a labeler, and a case packer. How are they doing?

Imagine we repeat the efficiency and throughput calculations for each piece of equipment. We won’t go through them in detail, but you can verify them using the formulas provided above. The table below shows the results for all five machines.

Machine MTR MTBF Max rate Efficiency Throughput
Decaser 1.5 min 30 min 240 bpm 95% 228 bpm
Filler 4.0 min 60 min 180 bpm 93% 167 bpm
Capsuler 30 sec 15 min 240 bpm 96% 230 bpm
Labeler 45 sec 10 min 240 bpm 92% 220 bpm
Case packer 3 min 40 min 300 bpm 93% 279 bpm


At first glance, this looks pretty good. It would be nice to see those low-90s move up a point or two, but surely we’re getting somewhere close to maximum throughput on this line, right?

Not even close.

Here’s why: Machines don’t all go down and then come back up at the same time. They all have different MTRs and MTBFs, and every time one machine goes down, it affects all of the other machines on the line. So, to find the efficiency of the entire line, we have to multiply the efficiencies of all of the machines together:

0.95 x 0.93 x 0.96 x 0.92 x 0.93 = 0.725

The efficiency of the entire line is only 72.5%.

You would never buy a single piece of equipment that was only up and running 72.5% of the time. If you did, you’d send it back to the manufacturer as defective and demand a refund.

But, wait, we’re not done. We still have to calculate the throughput.

For this, we use the individual throughput of slowest machine on the line. As they say, “a chain is only as strong as its weakest link.” On a production line, that means that the whole process can only go as fast as the slowest machine, which we call the constraint. In our example, the constraint is the filler, which has a maximum throughput of 180 bpm.

Total Throughput = Throughput of the Constraint x Total Efficiency
Total Throughput = 180 bpm x 0.725
Total Throughput = ~130 bpm

At this point, you’ve probably gone from thinking “That’s pretty good!” to thinking “Yuck.” Now you can see why so many lines underperform. This line has machines that can run anywhere from 180 to 300 bottles per minute, and it’s averaging 130. By running the line at such low capacity, we’re leaving a lot of money on the table.

How to maximize your throughput

You’ll be happy to hear that there’s a light at the end of this production line tunnel.

Since the slowest machine is the one that’s constraining production, this is where we need to focus. And don’t worry — it doesn’t require going out and buying a new filler! We can solve the problem by putting buffers in the form of accumulation tables before and after the filler. The buffers isolate the filler from the rest of the line so that even when other machines go down, the filler can keep on keeping on.

  • When a downstream machine, such as the capsuler, fails, the filler can create a store of products in the buffer that will be ready to proceed down the line as soon as the problem is resolved.
  • When an upstream machine, in this case the decaser, fails, the filler can keep working on the products that have accumulated in the buffer.

The buffers ensure that the filler performs at its maximum efficiency, rather than being affected by stoppages elsewhere along the line.

This simple solution is amazingly effective. Let’s do the math.

In the original example, the overall efficiency was so low because the individual machines were interdependent. If one went down, the whole line went down, and the total throughput was limited by the constraint.

When we add buffers before and after the filler, we effectively split the line into three sections:

  • Section 1: Decaser
  • Section 2: Filler
  • Section 3: Capsuler + labeler + case packer

The filler is still the constraint, but, thanks to the buffers, it can keep running regardless of what happens upstream or downstream. So, our new calculations look like this:

  • Decaser: 240 bpm x 0.95 = 228 bpm
  • Filler: 180 bpm x 0.93 = 167 bpm
  • Capsuler + labeler + case packer: 240 bpm* x (0.96 x 0.92 x 0.93**) = 197 bpm

*The max rate for this section of the line is the rate of the slowest machine.
**The efficiency for this section of the line is 83% (0.96 x 0.92 x 0.93 = 0.82).

Now for the kicker. Because the filler can keep operating even when the other machines go down and products can accumulate in the buffers, the throughput of the entire line is equal to the throughput of the filler. In other words, the throughput on this line is now 167 bpm, rather than 130 bpm. That’s an increase of more than 28%.

At this point, I hope you’ve gone from “Yuck” to “Wow!”

Here’s one more set of calculations to consider.

Before we added the buffers, the line was running at 130 bpm. At that rate, over an 8-hour shift, the total production would be:

130 bpm x 480 min = 62,400 bottles

With the buffers, that same 8-hour shift would produce:

167 bpm x 480 min = 80,160 bottles

Let’s assume that the winery makes $0.80 profit per bottle (a very low estimate). Running one 8-hour shift per day, 5 days a week, for 50 weeks a year, the additional profit would be:

$14,208 per shift and $71,040 per week, adding up to $3,552,000 per year!

You read that right. $3.5 million extra profit simply from adding two accumulation tables onto the line.

And just so you know, this isn’t a hypothetical exercise. This is a real example of a bottling line at one of our customers’ facilities. You may have heard of them. They make excellent wine.

If you’d like to learn more, here are several resources to help: