1 line checkouts are much better than multiple line checkouts – MBA Learnings

A few weeks ago, I wrote about why queues form. The one line answer is that they form because of statistical fluctuations and dependent events. The concept is simple – if your presence at a meeting is dependent on the previous meeting and the average time in the meeting is variable, it is likely that you’ll have people waiting for you, on average.

There’s a really cool application of this principle when it comes to checkout lines in stores and supermarkets. Multiple line checkouts are woefully inefficient.

So, if the supermarket next door replaced multiple checkout lines and replaced it with 1 line, it could reduce your waiting time to approximately 1/3rd your normal waiting time. Why? Because longer lines minimize variability. If you are stuck in a short queue with 2 coupon sharks who take forever to pay, your average waiting time becomes very long. Such variability is minimized in a single queue as it is unlikely you have a coupon shark at every checkout counter.

The beauty about 1 line queue systems is that it also feels fair. We all hate it when we see that other queue go much faster. The downside, however, is that single queues can look and “feel” really long. So, the conventional wisdom is to have multiple queues because long lines can turn off customers.

Whole Foods in Manhattan, however, decided to just ignore the conventional wisdom ten years ago and implement the more efficient single queue checkout. It has worked fantastically well for them. And, now you know why.

Whole Foods Manhattan(a line manager at Whole Foods Manhattan who makes sure people move quickly to the nearest open register)


Managing queues – MBA Learnings

I’ve been sharing a run of Operations learnings of late as a part of this series. This has been surprising as I never considered myself a fan of the subject. However, thanks to a combination of a Professor who’s more than managed to pique my interest and a realization that learning to manage business operations isn’t very different from managing life operations, I’ve enjoyed my time studying Operations. And, today’s topic is managing queues.

Managing queues is particularly interesting as we all experience, and generally dislike, queues. The average wait time in a queue is given by the following formula –

Queue formulaTo break down each of the parts of this formula in simple terms (with an analogy of a queue at an ice cream stand)-
– Mean service time is the amount of time taken by the person serving the queue
– Utilization is the amount of time he/she spends serving out of his/her total time on the job
– Variability (more on this later) is a measure of how steady the demand is. If people enter the queue steadily through the day, it is much easier to deal with demand versus random fluctuations

Things get interesting when we study the effect of utilization on increasing waiting time. This graph, from HBR, illustrates it beautifully –

What this graph is saying is that waiting times more than double when your solitary ice cream server is working 80% of the time. It doubles again when he finds himself working 90% of the time. Why? Because delays are caused by sudden fluctuations in the queue, e.g.,  a mass of tourists that come in to buy ice cream in the midst of their city tour. And, since our solitary server has no spare capacity, it is inevitable that waiting times go up.

So, if you are staffing your restaurant, for example, and if you find yourself running at 100% capacity all the time, that may not be the best thing for your customers since it is inevitable that waiting time goes up. (If you are an exclusive restaurant, it may not be a bad thing – that’s a different matter.)

The insight in managing our personal lives is pretty profound – if we organize ourselves such that we always find ourselves running at 100% capacity, it is inevitable that queues will build up on our plate. That’s because work doesn’t arrive at a constant rate. An emergency project is bound to show up and, if we’re running with no safety capacity, that could be a problem. Additionally, we’ll never have the bandwidth to deal with other sorts of fluctuations that may occur outside work – a family member that gets sick, a friend that needs help, etc. So, it is a good idea to maintain safety capacity.

And, how do you do that? Learn how to scope projects well. I had a manager who was a master at making sure we needed no late nights to get to the finish line on our projects. He believed our best work was done when we were relaxed. He reiterated that he’d rather we build models slowly, but accurately, rather than fast and requiring multiple revisions. He also believed we should always be able to deal with issues that come up with minimal stress. And, of course, he consciously developed this single skill that, in my opinion, distinguishes great managers from bad ones.

In short, he understood the importance of safety capacity. We should, too.