The ‘Happy Hour’ is a cultural institution that exists in bars all over the world. For certain explicit hours of the day, often in the early evening, a bar will offer drinks served at much lower prices than normal, the aforementioned “happy” hour(s). This institution of course has benefits for both customer and the bar itself. The customer obviously saves money and feels good about claiming a discount. The bar gets more customers into the premise than they otherwise would expect at these hours, which hopefully will attract more customers from the street and maybe even some of the Happy Hour crowd stay on for longer once the prices go back up. The exact drinks promotion that is offered during Happy Hour can vary greatly from bar to bar, with conservative establishments offering merely cocktails reduced by a Euro/Dollar/Pound or two, while others can offer free shot chasers with each drink. By far the most common Happy Hour however is the ‘two-for-one deal’, where when you purchase from a limited stock from the bar (usually just beer and certain wines), another of the same drink will be supplied free of charge. This simple Happy Hour rule is widely used all over the world, but it is far from an optimal strategy for the efficient distribution or intake of discount recreational alcohol. This article critiques the two-for-one Happy Hour deal, and proposes a modified variant which is shown to offer extra utility for both the customer and the proprietors of the bar itself.
The fundamental rule of a Happy Hour is that any income earned during those magical minutes is extra income: it is likely the bar would be practically empty without the incentive. The deal attracts customers who otherwise would not exist, based on the promise of reduced price alcohol. In the case of the two-for-one deal, the bar will always lose money on each transaction involving the drinks included in the deal. The hope is that the extra business flowing from the increased number of customers will still mean that the bar is making money at the margins, albeit very incrementally with each extra customer that is getting a half price beer. Ideally though, the bar management want everyone who gets a reduced price drink to still be on the premises after the Happy Hour ends, and willing to pay for a full price beer which would make up for the lost bar income. This hope is not entirely based in reality however, as two-for-one bars generally empty significantly once the deal ends, as all of a sudden the bar is just a bar, with nothing happy about it whatsoever.
On the customer side, the two-for-one deal seems to be a great deal. Going to the deal with a (precisely) even number of friends would mean that for the duration of Happy Hour, each would pay half price for a beer or wine of his/her choice. Not many two-for-one deals venture outside these cheap standard beverage options however, meaning consumer choice is severely limited. Once the deal ends, there is not much incentive to stay in the bar, as the customers simply came for the cheap drinks that do not exist anymore. If the customer is responsible and wanted to go home at a respectable hour, this is fine. Otherwise, it is either a premature end to the evening or the beginning of a search for a new bar. In summary, the two-for-one deal is costly to the bar owner, and but a throwaway gimmick for the customer. The model works, yet it could be so much more.
An Alternate Model, Incorporating Risk
For the duration of a Happy Hour, all drinks are normal price. A customer queues and orders at the bar, and when the staff bring his bill and drinks, an unbiased coin* is flipped. The customer calls heads or tails. If he is correct, his drinks are free. Otherwise, he pays. All drinks in the bar are available to order, yet he cannot order more than 10 individual drinks at a time.
( *A variant is to have the barstaff roll a numbered dice in a cup, which is emptied on the table. The customer guesses whether the face value is odd or even)
In this model, the incorporation of risk is argued to add a more satisfying Happy Hour experience for both customer and bar proprietor. Whereas in the two-for-one model, the bar loses money on drinks included in the deal at a probability of 1, in this variant the probability of loss in each individual order is reduced by half, to 0.5. The risk is therefore shared with the customer, and therefore the Happy Hour becomes a game of chance between the two parties. What is most interesting from the point of view of the bar proprietor is that the stakes at play are always defined by the riskiness of the customer. Someone who has won a few rounds of a coin toss may get cocky and overextend himself to ordering expensive cocktails or shots, and may end up paying the price. Of course, he is just as likely to get the whole lot for free, but this is the game. The bar will see this act repeated countless times, and the results will even out eventually to a point where the customers and bar share the cost equally. From the bars perspective, it appears to be not much better than the two-for-one deal, yet the difference is that in this model it is at least an opportunity to win money back, with varied returns due to the different stakes at play. In business, a 50% chance is much better than 0%, especially with the stakes raised. Further on, the model is also argued to offer a greater chance of post Happy Hour (and full price) customer retention than traditional deals.
On the customer side, the advantages of this model over the simple two-for-one are more subtle, yet immediately apparent. If it was not in human nature to get a rush from taking risks, then casinos would be empty, online poker would not involve money, and the international gambling industry would not exist. The simple act of placing a bet on the outcome of an event transforms the transaction into entertainment in itself. This is regardless of win or loss. Winning will produce momentary euphoria which the customer takes back to his friends, while losing will just want that customer to look forward to the next round, just for ‘one more chance’. With greater stakes, the risk of the customer increases, and these high stakes orders could even command the attention of the entire packed bar, just as in a casino. The introduction of risk, with its rewards of jubilation and also deflation, mean that there is raw human emotion flowing in a space where other Happy Hours just have people smiling over cheap alcohol. Here there is more mixing within the bar, more interaction with the staff due to their role in the game, and more camaraderie in the shared experience. I would argue here that all of this conspires to mean that there is a higher chance of customers remaining in the bar once the Happy Hour ends, perfectly content to pay full price. For as well as this hypothetical shared community, there is also the case of those who may have won a large share of their bets, and therefore have “extra”, discretionary income burning a hole in their pockets due to the natural high of winning a high stakes bet. All in all, a much richer experience than sharing the cost of a beer with a friend.
I cannot claim credit for the formulation of this model, for many readers (particularly those who were on Erasmus with me in Tilburg ’05) may recognise the basics from one place or another. I add merely a formal breakdown of the mechanics, as well as a few added constraints (the bar in Tilburg went bankrupt after a year) to prevent revenue losses for the bar reaching a critical level. This Risky Happy Hour has potential as not only a short-term gimmick, but also as an opportunity for a bar to show off its best drinks with the potential (from the customers point of view) of being completely free. Even without this, the bar will make a name for itself among many diverse crowds, from students to working professionals, meaning the opportunity for a much broader client base for the bar, and socialising experience for all involved. The bar may pay for half of the drinks, but due to the distribution of prices on order bets, that is not to say it pays half the bar bill. For a few hours, one night a week, they should take the risk.