As a customer, all you care about is latency: How long you spend waiting in line. As a store owner, you also care about throughput: How many cashiers you need to process some number of customers per unit time. A store owner that cares about customer satisfaction will still care about latency, of course.

Little's law doesn't imply what you said, either. It tells us the number of people in the system based on arrival and wait, but nothing about throughput.

And you're right about throughput, I was implicitly holding the number of people in the queue constant, and you can obviously add arbitrary delays and increase the number of people in queue without touching throughput.

Just like the website load times, you should be focused on the 99th percentile and the median. Single line reduces those metrics.

Really this boils down to whether the business is interested in saving a bit of money (fewer cashiers per purchase) or improving customer experience.

The metric you want is to figure out what wait time constitutes good/acceptable/bad service, and what percentage of people get that service.

Mathematically a single queue is far more efficient.

The psychology suggests that someone seeing a line makes them work faster.

First off, this is fleeting -- in a just a few years, we won't need people at the registers anymore as RFID technology gets cheap and more people become comfortable with the ever improving self checkout.

Secondly, even if you assume we will have checkers for a long time, this is again an easy problem to solve. You have a single queue, and then you make part of their pay based on average time per client or their evaluation. They'll work faster because of it (maybe).

A counterpoint to my own point is that Target does rate their checkers on efficiency, and I've watched them game it before. The checker meticulously lined up all my items with the barcodes in a row, then hit the "new customer" button, quickly scanned all the codes and hit the "total" button, giving him a very fast time. He figured out that they only measure from when he says "new customer" to total, so even though he was slower than everyone else, he had the fastest time.

But my overall point is that humans will soon be out of this loop anyway.

No. In the absence of behavioral effects such as the ones suggested here, the reason to have a single queue is not to reduce the mean time for a customer to check out, but to reduce the variance. With one queue per cashier, if you happen to get behind the person paying entirely with pennies, you're screwed. With a single queue, the additional latency introduced by such a customer will be distributed among all the customers who aren't being served yet, rather than just the few people behind them in the same queue.

Also, variance is what people perceive as bad service, not the mean time. If everyone takes a little longer they feel happier because of the perception of fairness.

I'd say a large bottle neck here is the employee base. If they factored in throughout into bonuses then they could get rid of many employees AND look good while doing it.

"New incentive to allow workers to make more then minimum wake"

I could already imagine the native ads!

I was very confident this idea would give better outcomes as I knew potential workers who could easily handle the higher load and do a better job at that.

Regrettably my idea was not used, and true to form, the leisurely pace of work was a drag on productivity. Twenty years later, long after I'd left that workplace, the outfit was essentially bankrupt and called out for mishandling funds.

I still think I was right, but even good ideas can be hard to sell.

Their model is probably less efficient during regular operation, but the variance when problems arise also seems lower.

As it was expressed to me the basis for the decision was that the "leaders" didn't believe the workers could handle the load as I suggested, so they went with their original plan.

At the time I attributed it to their rigidity but it could have been about relative salaries. Here I thought doing it my way would actually cost less when all added up.

It sounds like a trap. Move quickly, get higher scores for that, yet get reprimanded for poor customer service.

From experience, some customers are simply difficult to get to move on. You can do some things, but go too far and you wind up getting complaints about rudeness. It is all a balancing act.

Regular slow customers you can call for backup, if the place allows for that, but sometimes you are simply stuck.

This problem has improved significantly in the past couple of years, I imagine due to Angela Ahrendts takeover of the stores, but I still find it annoying.

For whoever's downvoting: for starters, no paper receipts, no bags, no cash. With self-checkout that's all my choice. Not so with Apple checkout.

The self-checkouts are mostly for basket-only shoppers.

It would have been interesting if they'd thrown the self-checkout option into the mix. I suspect people self-scan much slower than a cashier, but given that self-checkouts occupy less space than a full till with conveyor belt, (hence a lot more terminals), I suspect overall they're still faster than using that space for normal tills.

The machines also have a pressure sensor that detects that you have placed items on it after scanning, if you don't allow time for the sensor to settle the machine will give an error and require a cashier to reset it.

Interestingly the same scanners can be used for the online order picking.

Used to be that you could make small talk and feel the hole in your pocket growing bigger while you handed over that old crinkly dollar bill. Now I can hardly get out a perfunctory "Hello, how are you doing?", because I'm too busy trying to figure out whether I need to explicitly choose credit or whether I'm going to need to sign or not given the dollar amount... I'm not suggesting that we go back to the "good old days". I would, however, suggest that we progress in the direction of pathos and our shared humanity rather than toward whatever it is that tickles the fancy of Homo Economicus.

> Used to be that you could make small talk

If you want to satisfy as many customers as possible, you'd need to handle both: the customers who want a human face, and the "is there a machine I can speak to instead" customers. Not everyone considers human interaction a feature in a transactional setting.

Economics in action. Lots of people. You don't have to engage with the small talk if you don't want to.

I'm in Austin, TX. Looks like you in the UK. I'm going to look into whether there is something like this locally. Thx!

Also, just curious. What are your thoughts on the EU vote coming up?

My thoughts?

1) Look at a map centred on 52N02W with a radius of 2000 miles and think about what the future is for an island off the mainland with a population about two and a half times that of Texas.

2) Daniel Kahneman would have a field day over here at the moment! A fair spread of his hypotheses can be seen in action daily.

PS: I was describing Birmingham's central food markets in Digbeth by the way, but other cities have good markets as well - open three or four days a week and selling the same kind of stuff you can get in supermarkets without the packaging. We also have more expensive/artisanal 'farmer's markets' with higher priced produce in middle class areas, typically one Saturday a month.

plus, this appears to be the result of an experimental setup rather than real-world facts. It's trivial to get real-world data from some place that implemented the difference, since all receipts are time-stamped. You just have to count the number of receipts in single-line, vs everyone-with-their-own-line, and see which is bigger. (You can also see how average total or number of items differs between setups, or normalize for any changes, such as only considering the time taken to ring up literally the exact same order.)

this isn't data rocket science. any store which went single-line should have this data.

But doing it in a lab means you have no ecological validity. People know they're being watched closely, much more closely than they would be in a real life scenario, and that will affect their behavior. So you really can't be sure your results will generalize to the real world.

Possibly the best you can do is take a few stabs at both approaches and hope that all the studies point in roughly the same direction.

I hit all of them every time I go. It's awful. The old person returning a head of broccoli, the kid pulling down a display stand into my shopping on conveyer, the milk bottle leaks onto cashier. Every time I go it's a new kind of pain.

The fact that the outcome will vary in different store environments isn't a trivial factor in evaluating the claims of the experiment. IOW the findings may apply in some cases but not others, which is an important kind of detail to consider.

As a customer I only, only care about my perception of speed, not actual speed.

Multiple lines 'seem' fair, there's also little variance and hence I leave feeling positive most of the time.

A store might save a small percent on cashers using this research but pissing the customers off the huge amount it would seems like poor economics.

Which in turn can be gamed. See the infamous story from Houston, where there were large numbers of airport complaints saying it took too long for arriving baggage to reach the bag claim. The solution was:

http://www.nytimes.com/2012/08/19/opinion/sunday/why-waiting...

(they rearranged things to increase the distance between arrival gates and bag claim; it still took the same amount of time to deliver bags to bag claim, but now passengers spent more time getting to the bag claim and less time waiting at bag claim, so they perceived it as faster and complained less)

From my airport experience, hybrid models tend to be pretty good: one single line to security scan or passport check and then it is ushered into many short ones. Probably should mathematically model that and see what happens.

The notification comes when the cashier moves to the payment step, so you still have time to unload the purchases onto the belt before the previous customer is finished.

There is a very real stress you feel when a line starts building up. I naturally worked harder when it felt like I was falling behind.

  First they came for the waiting room, and I did not speak out -
  Because I did not mind to queue.

  Then they came for the stock pickers, and I did not speak out -
  Because I did not mind to pick.

  Then they came for the cashiers, and I did not speak out -
  Because I did not mind to be anti-social.

  Then they came for me - and there were only robots left to speak for me.

I can't believe they've ommited a situation where this one lonely customer answers "Yes, I DO want to hear more about the offer" and starts signing up for the store membership, a bonus program, a store credit card, a cat insurance, and a dunce hat.

For example, at a hotel checkin, flight baggage checking, or Fry's electronics, there is one line which feeds into multiple agents. As the number of agents grows, there's a nontrivial amount of time where an agent is done with the current customer, but the next person in line has to walk maybe 50 or so feet to them, wasting some time, especially if they are slow (kids, luggage, elderly).

APU: "Look, Mrs. Simpson, the express line is the fastest line not always. That old man up front, he is staved for attention, he will talk the cashier's head off."

I think The Simpsons got it twenty years ago: when it comes to checkout lines, the human factor means far more than the mathematics.