Constraint vs. bottleneck

In Theory of Constraints lingo, there is a subtle difference between a constraint and a bottleneck.

A bottleneck (resource) is a resource with capacity less or equal to demand while a constraint is a limiting factor to organization’s performance, an obstacle to the organization achieving its goal.

A constraint can be called bottleneck but a bottleneck is not always a constraint.

>watch this article as video on my YouTube channel

Watch the video

Watch the video

Let’s take an example of a plant with a subassembly workshop gathering resources A, B and C. The whole process needs another resource D and final assembly consisting of resources E and F. The capacity of each resource is displayed under their letter.

The demand is 100 units per day.

According to definitions we’ll find two bottlenecks: resource B limited to 80 units/day and resource E limited to 60 units/day. Each of these two have a capacity less than daily demand.

Resource B is handicap to resource C and for the whole subassembly workshop, but has little influence on the throughput of the plant. Plant’s throughput is limited by resource E, which is both a bottleneck and the constraint. It is primarily E which hinders the plant to deliver 100 units/day.

Imagine The subassembly is led by a foreman named Hector. Hector’s realm encompasses The resources A, B and C. The final assembly process is his customer.

Hector has significant experience within this company and is well aware B is a bottleneck. Even so Hector may not know anything about Theory of Constraints, his common sense made him discover some good rules to better exploit the bottleneck resource.

For example, Hector organized breaks so that B is never left unmanned and not running, manages to minimize changeovers.

If he knew about Theory of Constraints, he would probably squeeze more throughput from B, for instance placing the quality check before the bottleneck in order to insure only OK parts will be processed by the very limited B. Actually quality check is after C, which sometimes causes B to waste valuable time processing parts that will not pass the quality check, something that could be foreseen before B.

As it is the case in many companies, top management set local productivity objectives and is expecting Hector’s subassembly to run with best productivity. Logically Hector will complain about B’s limitations and keep asking for another investment in a second B. Waiting for this investment, Hector manages to produce daily around 80 units, the best subassembly can do.

In Hector’s eyes B is the constraint, which is true if we consider subassembly alone.

Production manager Isadora has to take care about the whole plant and thus considers the whole process. She doesn’t know either about Theory of Constraints, but her analytical skills and common sense focused her attention onto E, the bottleneck and constraint to the whole process.

Having limited means, she’ll explain Hector that working to increase the capacity of B would have little interest as long as E is the limiting factor for the whole system (the plant). What Isodora did not notice is that as long the daily limit is 60 units/day, some costs could be saved in subassembly if its daily production would be aligned to the capacity of E, for instance overtime and excess inventory carry over costs. But she’s blinded by local productivity objectives set by top management.

Nevertheless, Isadora came close to self-discover the five focusing steps of Theory of Constraints:

  1. Identify the constraint (E)
  2. Exploit the constraint
  3. Subordinate everything to the constraint (e.g. subassembly)
  4. Elevate the constraint
  5. Prevent inertia to become the constraint

If Isadora succeeds to elevate the constraint E, chances are that the B will be the next constraint!

Related: How to identify the constraint?

About the Author, Chris HOHMANN

About the Author, Chris HOHMANN

View Christian HOHMANN's profile on LinkedIn

16 thoughts on “Constraint vs. bottleneck

  1. In the above illustration E is bottleneck as well as a constraint. If we want to speak of Bottleneck Vs Constraint – we should make the illustration as – B (60 or less than daily demand throughput of 100) and E (80 or 90 but greater than throughput of B). In this case we are in a dilemma of Bottleneck (B) Vs Constraint (E).

    Am I right Chris? It would be good if you provide your inputs on resolving the situation I shared in above para.


  2. I have always felt that TOC has always missed the main constraint on the performance of any organisation. If I remember correctly, Eli Goldratt almost recognised this in ‘The Goal’. The fact that Herbie was the constraint on the pace of the troop was used to clarify the concept of constraints on a group activity. What he didn’t identify was that the inability of an organisation to release and focus the ability of all their people is the biggest constraint on their performance, and a self inflicted bottleneck. Removing this bottleneck, and releasing the flow of all your people’s abilities*, is the key to reducing all the other constraints and bottlenecks.
    * Ability has three dimensions.
    Talent. The ability to do existing tasks well.
    Creativity. The ability to improve what we do and the way we do it.
    Enthusiasm. The emotional ability/energy to do the first two.


  3. Hello Chris,

    The above article defines bottleneck as a resource with capacity less or EQUAL to demand.

    1) It is not clear to me as to why we should call a resource a bottleneck if it is able to meet (equal) the demand.

    2) Some authors define bottleneck as “resource that requires the longest time in operations of the supply chain for certain demand”. According to this definition even if a resource throughput exceeds demand, it can still be considered bottleneck if it takes the longest time in operations compared to other resources. This definition is different from the definition you have mentioned in the blog. I appreciate your views on this matter.


    • Hello Anil,

      Thank you for taking time to share your thoughts and give me opportunity to check clarity and consistency of my post.

      1) The bottleneck definition as stated in my post is found in Philip Marris’ book (French) as well as in the TOCICO dictionary, 2nd edition 2012 for example.
      When the bottleneck capacity is just equal to the demand, it is unlikely to exploit it 100% and any event hindering the bottleneck to run 100% will end not fulfilling the demand.

      2) If a resource capacity exceeds the demand but is the slowest or longest step in the process, no energy should be wasted on it as the constraint is elsewhere, probably in sales, marketing or wherever.
      So yes, you may call it a bottleneck if you like, but should not work to increase its capacity as long as the demand can be fulfilled.

      Hope this clarifies a bit?

      Kind Regards,



  4. Hello Christian,
    Please clear some doubts I have on a multiple choice question:
    Here is the question and choice of answers….
    A bottleneck is:
    a. a work center that has the highest or least number of jobs flowing through it as in either case it will starve the nearby work centers
    b. a job that involves complexities and hence requires longest time to process
    c. a work center that has less capacity than prior and following work centers
    d. a job that is simple and hence flows through the system the fastest thus starving the production centers for “want of more”
    Here are my doubts:

    1) In option “a”, how can least number of jobs flowing through it can starve the nearby work centers ? In case of highest number of jobs flowing through a work center , I believe other work centers will not comparatively have much jobs to do and thus starve.

    2) I am referring to option “b”. I am not choosing this as right answer as this option is talking about “a job” and not “work center”. If it were “a work center” that involves complexities and hence requires longest time to process , I would consider this as possible answer. Is my understanding correct?

    3) I am referring option “c”. Can I call a work center that has less capacity than prior and following work centers a bottleneck? If yes, I will have three definitions of bottleneck namely –

    1) A bottleneck (resource) is a resource with capacity less or equal to demand – This is what your initial post says.

    2) bottleneck is a resource that requires the longest time in operations of the supply chain for certain demand – This is what was explained by you in response to my earlier query.

    3) a work center that has less capacity than prior and following work centers – This is the right answer given for the above multiple choice question.

    Now I have three definitions of “bottleneck” which are all correct. The problem I am facing is – in a process the bottlenecks can be different to each person depending on which definition he or she is following. This could be quite confusing and this also requires one to define bottleneck before resolving the issue of bottleneck in an organization. Do organizations really practically do this (define bottleneck) before resolving the bottleneck?

    Has the developer of Theory of constraint (TOC) – Eliyahu Goldratt defined bottleneck in his theory ? If yes, please let me know how he has defined it.

    This is a very long post. I thank you for your patient reading. I await your response.




    • Anil, I am sorry to have to keep my answer short: the only bottleneck that is worth caring about is the one limiting the Throughput of the whole system.
      It is then called “constraint”. Other bottlenecks may exist in the system (the topic of my post) but we have to care only about the constraint and watch others would-be constraints in a preventive risk mitigation.

      Further explanation would take us a bit far and frankly speaking I cannot spend more time on it for the time being.
      Sorry if I let you half way across the river.



  5. Christian- A slightly more realistic discussion (and longer) would include things like “efficiency” measures that the managers are asked to follow. This would push them AWAY from focusing on B and E, but on the work stations, like F, that have chronically low efficiency. They might even be induced into pulling in more jobs to load up F (and A, C and D) to get their efficiencies up. Of course, this will likely kill overall throughput. (I have the voice of the shop manager from The Goal movie running through my head, “But the one thing you’ve got to admit is that our efficiencies are up.” — While the plan is bleeding money.)


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.