When a day like 31st December happens to be a Sunday, you are very unlucky if living in Austria. Grocery stores are usually closed Sundays with only a few exceptions. Those exceptions already tend to be crowded on ordinary Sundays, and even more so if it is New Year’s Eve. All the people who didn’t do their shopping for that day and the following holiday on time are then blocking those few grocery stores that are open. So I too was part of a jam in front of the only „Billa“ (a main grocery chain in Austria) that was open on the afternoon of the 31st. Unexpectedly, this holiday madness became a Eureka-moment for me. Because as I told you: I was standing in the jam in front of the store.
Let’s think a few moments about that: If you would let in all these people at once, who are hungry for salmon, thirsty for sparkling wine and killing for fresh baguette – what would happen? Each of them would have a rather shitty shopping experience. None of them could move on as they would like to, you would be shoved around and maybe you wouldn’t even get what you were looking for because the employees wouldn’t be able to refill the shelves on time and suppliers couldn’t get through – classical bottlenecks so to say. Maybe you wouldn’t even try to get everything you need because you would just want to get out again as soon as possible. But even then it would take you a long time to get out if you weren’t one of the lucky ones who were swept into the store at the head of the wave. So in any case it’s highly probable that you would leave without having all that you intended to buy.
The sight of a hysterical crowd would maybe also deter those who are lucky enough to be still outside the store from moving any further because they want to avoid the chaos. So the store would lose some substantial amount of business and altogether it would be nothing but stress for customers and employees alike. Unfortunately, I see situations like this in many organizations every day: Especially in knowledge work companies the gates are never closed because management still believes that people have to be occupied a 100 % all the time. So things like quality und what’s of value to the customers falls by the wayside. If people are pushed into a grocery store uncontrolled – that cannot turn out well. Neither will more work be finished nor will it be finished any faster if you push more and more work into the work system.

Inflow regulation by Billa

The responsible Billa-folks must have thought the same. In front of the main entrance they positioned security who would close the roller shutter whenever they got an „Enough“-signal from the inside. Only when a sufficient amount of people had left the store through the side entrance, the security personnel would open the gates again for the next group of people. So the people inside the store were able to do their shopping rather relaxed and quickly and the system didn’t break down because the shelves were always refilled on time. The people in front of the store could see that the queue was moving forward at regular intervals. So everyone was still able to decide wether they wanted to wait or not. Those who still wanted to enter the store could roughly estimate how long it would take them to get in. In this way, Billa optimized the throughput and I guess that they earned more money by doing so than if they hadn’t taken these measures.
That’s also the way in which we should operate our work systems. We should consider: What’s the optimal amount of work in our system? „Optimal“ means to be occupied well but still the work moves forward smoothly and we have enough capacity for adressing problems if they arise. Through limitation it becomes more predictable when work will be finished and the next work can be started. So we should only start as much work as the system is able to bear without getting out of balance – all new work therefore has to wait outside of the system first. That’s why I like the alternative term for WiP-limits coined by Prateek Singh: „Optimal Operating Capacity“. Or shall we just call it the Optimal Amount of Work in the System?