I came to the review page expecting a rating and a review that I would have to write under only to find that I evidently read this before I made rating and reviewing a mandatory step in my reading process.
There are a few takeaways from this on a second read and keeping in mind that this represents novel developments from the other works in existence. Goldratt has continued the process of ongoing learning in his own work. We got the other TOC books (not Critical Chain, etc but the ones that revisit the same work in The Goal) because he realized that while the narrative was necessary to keep people reading it, it was insufficient at teaching the method. In observing the case studies in application of the ideas from all of the TOC subdisciplines, the discovery made was that the application of the ideas isn't the problem. People applying the ides have generated a bank of solutions and made the entire landscape of problems in the various context-appropriate areas (production even if it's not called production, project even if it's not called project, etc) visible and simple. The problems are that people aren't applying the ideas and that when applied, will not see the results that they can possibly achieve due to adherence to rules created based on the bottlenecks being resolved by the thinking processes. Goldratt also elucidates that the terminology is different in different industrial contexts but the processes, and thus the UDEs, are exactly the same. In medicine, they have a production model but the WIP is people, and bottlenecks don't create scrap. This invites us to consider the larger applications of TOC outside of the areas already explicitly covered in the books, including the Handbook (which only includes a limited amount of information on other fields like Education). Importantly, he reinforces some central points from TOC and the theoretical basis for everything. The first of which is that the fundamental aspect of TOC is that it treats what we view as multivariate multi-problem environments as cases of mistaken analysis. Through TOC, we view these as problems that only appear to be distinct and have different causes because of our limited interpretations. Breaking the cloud is a systematic way of breaking that assumption and arriving at the core problem, the solution of which resolves all derivatives. This leads us directly to one of the main conclusions of TOC, which is that optimizing anywhere that isn't the bottleneck is fundamentally counterproductive and not a solution at all. If it doesn't serve the biggest goal, which for a company is always 'to make money now and in the future', then it is not really useful. If it doesn't impact the bottleneck, we're just exacerbating the UDEs. Goldratt points out that this entire perspective is a first-principles derivation from understandings of certain core concepts in physics, as opposed to other disciplines, such as problem, complexity, and the contradiction. Complexity is defined differently in that physics measures complexity by the number of levers you need to push to impact the entire system. Problem is defined as an unresolved conflict (as explicitly juxtaposed with 'something you don't like'). Contradiction is defined as a non-existent issue of analysis. There are no contradictions from the TOC view (or the physics view) but rather errors in parsing reality, poor assumptions. One of the things Goldratt is funny about, and this is ironic considering his stance on the social sciences, is that he has a ton of unstated assumptions/ conclusions about humans. For example, he says that metrics necessarily drive behavior. While this is true, he doesn't tie it in as a first-principles derivation. He just leaves it as an observation of organizational psych (somewhat forgivable considering he's saying he won't go over concepts that have a wealth of work on them already, like Deming). In any event, between those differences in ideas/ terminology, his unstated assumptions about metrics and systems impacting human performance and his contention that everyone is used to working using evolutionary (incremental) processes rather than revolutionary (big lever, big impact) processes, all of the various thinking processes and tools can be directly derived. Goldratt also goes through lengths to identify archetypal obstacles to the change process. He identifies 2 main people. The first is the impatient charismatic leader who can derail the change process, which is necessarily granular and lengthy. The second is the conservative who has lived through other change processes and views them as ultimately a waste of time. Goldratt proposes covering broad swaths of interesting material to keep the first kind engaged and making sure to identify the core problem, make sure everyone is clear on how this change will solve it, and to reach true consensus (which takes the form of everyone vocally working through why they think the core problem is the core problem and how it manifests in their area). This consensus creates the necessary mindset of both group cooperation towards the same shared objective for the company and individual area ownership. On metrics, Goldratt specifies that using a bunch of metrics is poor management for a few reasons. The first is that it sets up impossible standards against which an employee has to make judgement calls about which standard to fail to meet, rather than having the ability to meet the few true standards upon which their production should be appraised, which also creates opportunities for management to be predatory in reviewing employees. The second is that it motivates poor performance. Rather than working on what they actually need to, these metrics can force inefficient work habits. For example, measuring people based on man-hours causes people to take as long as possible on projects. This is a similar effect to what happens if you convert estimates for completion into due dates. People's behavior negatively impacts the project resources/ timeline/ status in an attempt to meet deadlines which can impact their job security. In discussing this, Goldratt reveals one of his heuristics for acceptance of an idea, which is that it is simple and pervasive enough to already have a word associated with it. The 2 measures that Goldratt advocates using in evaluating employee performance to goal fit this description, and are whether you did things that you weren't supposed to do, Effectiveness, and whether you failed to do things you were supposed to do, Reliability. Wasting time, energy and other resources doing things that do not contribute to the goal makes you ineffective. Failing to do the things you're supposed to do makes you unreliable. Both make things more difficult for you and others you work with and can cost business and relationships. These are the 2 metrics that employees already use when evaluating which strategy to employ in the face of a constrained situation. For example, in the case of estimates converted to deadlines, employees see that their reliability can be called into question if they state their most probable deadline and fail to meet it but that their effectiveness will be questioned if they give themselves enough buffer to definitely finish the work but then get it back to the requester earlier than expected. So, the behavior this set of constraints generates is that people give the maximum allocation of buffered time to their section of a project and then use every bit of it to appear maximally effective and reliable.
Where I take issue, and this is probably just my unfamiliarity with applications, is that he says that the process needs to be an 8 day process to work through each of the tools, from the current reality tree to breaking the cloud to implementation, and identifies the most important part of the process at happening outside of the conference space and when they go get drinks afterward. That strikes me as basically wrong, but I think that maybe it's because I'm failing to account for Goldratt's observations on the processing time each individual requires apart from the group. I just feel strongly that we can do better than that and due to the fact that he didn't do a first-principles derivation on the subject, which is entirely possible, I believe we can build on the work in that weak spot.