Wrong: improvements made anywhere besides the bottleneck are an illusion.
There's a quote about TOC I've seen quoted a few times recently, which seems to sum up the essence of TOC quite well.
It's misleading and wrong.
It's this: "Any improvements made anywhere besides the bottleneck are an illusion."*
Let me give you an example of a situation where it's not true.
This happened today.
My wife, daughters and I, went for lunch at a newly opened farm-shop restaurant.
- We arrived at 1:25pm, joined a small queue then got a table quite quickly.
- We were seated at 1:30pm.
- We placed our order at 1:45pm.
- Our food arrived at 2:30pm.
- We left at 2:45pm.
Yep: 75 minutes for lunch, 10 of which was spent eating, the rest waiting ....
The restaurant's bottleneck was, clearly, the kitchen.
And, yes, if there was any way to easily improve their throughput then that'd be very worthwhile, but you really shouldn't STOP there.
In fact, you probably shouldn't even START there. You should start by finding the kitchen staff a little thinking / breathing space so they can improve.
Maybe the easiest way to improve the situation would have been to remove or hide some of the tables (perhaps by putting "Reserved" signs on them?, perhaps by collapsing them and moving them into storage for a few weeks), then politely turn people away, "I am sooooo sorry, but we are full.")
That would have instantly reduced the pressure on the kitchen staff and let them think about how to better organise themselves. You know how hard it is to think about improving stuff when you're drinking from a fire-hose. They would seated fewer customer but served more, probably up-sold a few desserts & coffees and increased their revenues, and they definitely would have gotten fewer negative trip-advisor reviews.
Other improvements they could have made which didn't directly touch the kitchen? They could have raised their prices - over all, or just on the freshly cooked stuff. They could have changed the menu and specials board so it required less kitchen time during peak hours. My wife noticed they'd run out of soup, for instance - that shouldn't ever happen in a restaurant coz it's so cheap and easy to make in advance and loads of people order it AND perhaps the waiting staff could dish that up without needing a chef? So, tomorrow, cook more soup.
Are there other non-bottleneck improvements? Perhaps the kitchen's precious capacity is wasted because the waiting staff (who were learning) were taking multiple orders before dropping them to the kitchen, causing the kitchen to have unnecessary downtime? Perhaps the waiting staff were slow taking the plated food to the customers, which cluttered the limited kitchen space and slowed the chefs down? I wonder how much precious kitchen time was wasted when customers left out of frustration, even though 3 or their 4 dishes were cooked?
And, what if you discovered your waiting staff were making a lot of mistakes? (Our bill was wrong - we were overcharged but could easily have been undercharged). Wouldn't you fix that first?
Some improvements to non-bottlnecks are detrimental, some are probably "illusions", but some are vital.
Yes, bottlenecks let you focus, but you still need to think of the system as a whole.
[Update: I'm assuming that people will read this to mean they should only change the bottleneck. That's how I read it for the first few years after I discovered TOC, based on my interpretation of The Goal, before I discovered strategic bottlenecks and the meaning of subordination and buffers and ropes.]