Let’s say that you are the lucky owner of a prosperous bar called “Frank’s Beer and Daiquiri” that sells only two types of drinks, namely beer and (you guessed it) strawberry daiquiris. Interestingly, you have a crazy bartender called Frank who insists on deciding what people should drink. A customer will approach the bar and ask for a beer, and Frank will tell him, “No! You vill be having ze daiquiri, or else!” Besides having seen to many World War Two movies, Frank is a bit of a prima donna, but for some reason, your customers love it, and they actually drink whatever he orders them to.
Now, you are interested in maximising your revenue from the bar, so you examine which of the two drinks you make the most money on (and hence, which drinks you should ask Frank to ‘recommend’ to people). A brief calculation tells you the following:
- A bottle of beer costs you three dollars to buy from the wholesaler, and you can sell it to your customers for four dollars. Ergo, you earn a dollar every time you sell a beer.
- The ingredients for a strawberry daiquiri are more expensive, costing you six dollars. But the great news is that you can sell it for eight dollars, making you two dollars every time you sell a daiquiri.
From this, you draw the conclusion that Frank should be recommending the daiquiris to everybody. Right?
Well, maybe not. Because once you take a second look at the issue, you discover something else: making a strawberry daiquiri is considerably more time-consuming for Frank than popping open a beer bottle. This is no problem when Frank has lots of time, say, in the beginning of the evening. But when the bar begins to fill up, the picture changes, because now, there is a queue at the bar: Frank’s time has suddenly become the bottleneck, limiting how much you sell (Frank, being a sociopath, will not let other bartenders help him). And by observing the bar carefully, you discover that Frank can sell three beers in the time it takes him to sell one daiquiri. In other words, it seems that you are better off having Frank sell nothing but beer.
So far, so good. But then, another issue comes into play. It turns out that if your female customers are offered nothing but beer, each of them drink fewer drinks over the course of the entire evening. There’s just something about beer that makes them feel bloated and ready to go home after four bottles. And once the girls leave, it is only a matter of time before the guys take off as well, preferably after slapping each other around for a bit. As to the daiquiris, however, the female customers actually think the strawberry is healthy for them, so they guzzle them like no tomorrow and stay up all night, for a total sale of eight daiquiris per customer. Acknowledging this, you once again order Frank to sell only daiquiris. This finally maximises your revenue from the bar, and all seems well.
But of course, a new problem arises: since you no longer serve beer, all the ruggedly handsome coal miners that used to frequent your bar will disappear in search of a less posh place to chug down their after-mining pints. And to your chagrin, you discover that the wealthy and attractive thirty-something women who used to frequent your bar only did so because they secretly fancied the manly coal miners. These women leave as well, and suddenly you are left with nothing but effeminate philosophy students and the teenaged girls that mysteriously fancy them, and neither can afford to drink anything but mineral water, since they spend all their money on mind-altering drugs. Tough luck. You decide to fire Frank and sell drugs instead, starting an illustrious life of crime and passing outside the remit of this blog’s subject.
Now, when I made you suffer through this story, it is because I want to illustrate three things.
- First, detailed numerical analysis can be helpful. It can help you get a precise understanding of the cost structure of your bar, and this can in turn help you make more money.
- Second, numerical analysis should complement, not substitute, your common sense. You need to consider the numbers alongside more intangible factors like your bar’s atmosphere. Steering by the numbers alone is a Titanic kind of move, only without the Oscar nominations.
- Third, try to experiment and notice what happens. Bars and nightclubs are complex systems, and it can be difficult to predict ‘from the armchair’ how to improve them. In such situations, it can be a good idea to run little experiments and look closely at the results. For instance, what would happen if you rearranged the order of the drinks on the menu, or wrote some of them in larger fonts? Could you influence what people ordered? My guess is yes; for every Cosmopolitan aficionado out there, there’s probably a guy who just orders one of the three topmost items on the menu. There are hundreds of small things like this that you can safely and easily play around with, and that could turn out to make a real and valuable difference.
In general, as I have written about in one of my other blogs, Fragments of Knowledge, it is very often the case that the deity is in the detail: for more on this, check out the post called The Death of Complexity and the Rise of Small Things.