Experimenting with milkshakes?

I have been on a mission to convince firms to do simple experiments that will give them feedback regarding the decisions that they make. Just as with people (as Anders Ericsson studies), firms cannot learn without feedback. It turns out, however, that it is not easy for people in companies to see the wisdom in experiments.

Which is why I was so heartened the other day when I was with my wife in a restaurant, and I noticed that our two menus seemed to be identical except for one item. On my menu it was called something like “Coffee mocha milkshake” and on my wife’s menu it was “Chocolate Kona milkshake swirl.” The names were better than that, but one stressed the coffee part of the milkshake and the other stressed the chocolate part.

I hadn’t thought of menu descriptions as something restaurants should experiment with, but when I saw this, I suddenly realized it was a good idea. Do the names you give the food matter for what gets purchased? Do the descriptions affect the total amount of money spent, the satisfaction with the meal, or the likelihood that a diner returns? Can experimenting with the descriptions help the chef to know what types of food people do and do not like?

Fascinating, I thought, that this little restaurant would be experimenting with its menu (although if I were doing the experiment I wouldn’t be giving different menus to two people in the same party, but rather, to different parties). Perfect fodder for our next book! When the waitress came over, I asked her about the two different menus. “I’m so sorry,” she said. “I must have given you one of our old menus by accident. Let me take that one and throw it away and get you a real one.”

Oh, well, I guess that little restaurant won’t be an example in our next book after all. But the choco/mocha/kona milkshake was quite good my wife tells me.


You might be interested in this little diner here at Purdue in West Lafayette, IN.

They have two menus, one orange one yellow and while the menus have exactly the same items on them, they have very different prices.

The yellow menu is for normal times, the regular diners, and the Orange menu is for the "bar hours" the times when the bars are letting out in this college town, and the diner is full of drunk college students.

Can you guess which one has the higher prices?
Can you guess which


I know that some restaurants have had success with adding some high-price dishes to their menues - not because they will be sold but because they look the rest of the menu look cheaper.

It seems quite obvious that there is a whole lot of experimentation to be done with these menus - although I hope you're not planning to write a whole book about it... :-)


One interesting behavioral economics experiment I've always thought would be interesting would be to observe or survey people waiting in lines to board Southwest airlines flights. It's always been peculiar to me to see people hunkered down in line an hour early for a flight, particularly if the flight itself is under an hour. Why stand or sit on the ground for an hour before the flight just to have the possibility of a "good" seat for such a short flight?? This behavior is even more bizzare when there's a flight delay. For example, last summer I was on a 50 minute Southwest flight from Ohio to DC, and they announced a two hour delay. Yet, instead of sitting down in a comfortable chair in the lobby, most people stayed in line, and had to stand or sit on the ground for two hours! All for a good seat in the airplane for 50 minutes? Weird.



Maybe you came into the restaurant 3 months after the experiment actually took place. Don't write them off just yet!


This reminds me of the expression "menu cost." :-(


"There could be a prediction market that restaurant employees participate in. The contract would be a conditional one on whether 'coffee mocha milkshake' or 'chocolate Kona milkshake swirl' nomenclature generates the largest amount of sales. This would save on menu costs."

This is a completely inane idea.


In my area a ritzy shopping mall opened. One of the Italian places was failing and couldn't figure it out. Eventually the owner realized that people were looking for "classy" places. He raised the prices and didn't change anything else, now it's a success.

Anonymous Coward

Have you read _Mindless Eating_ ? Lots in there about things like menu descriptions, music, colors, etc affecting what people eat, how much, how long they stay, etc.


I'm surprised you did not reference Brian Wansink's Mindless Eating - also at mindlesseating.org. (FYI, the book describes research into eating and food choices. It is not a traditional diet book.) Researchers have played with menu descriptions in their research on how people eat and how they perceive value. For example, they've served the same wine with different labels and found that people believed the better labeled wine was better. They found that serving the same food in a more attractive ambience dramatically affected perceptions of the food.


I thought it had something to do with gender. Give the guy the menu with the straightforward wording, and give the woman the menu with the drink that emphasizes the chocolate and has more "flowery" wording. It makes sense, actually. Do guys really feel comfortable ordering "swirls"?

Tom Thorvaldsen

Many years ago, I went as a tourist to a delicatessen in New York City. I think it was the Stage Door Deli. After I had ordered, I noticed the regulars were ordering cheaper (but still humongous) meals. The menu has pretty pictures for the expensive tourist meals, and regular print for the cheaper meals for the locals.

The Cottage Economist

I think that Steven has a good point here. Too often, we try to bypass the actual information gathering pahse of a decision in the interest of saving time. But it is nearly impossible to actually know what a customer will choose without giving them the opportunity to actually make a choice.

The Cottage Economist


This is funny Steven, because I'm running a similar experiment with decision feedback in the dining setting with this project: http://thesecomefromtrees.blogspost.com

The idea being that diners mindlessly, and unintentionally, over-consume paper towels and napkins in restaurant settings, and that a quick reminder to them of the true cost of it will effect consumption at the margin.

In our limited pilot at a coffee shop, just remind people that paper towels come from trees with a sticker on the paper towel dispenser that says "Remember...These come from trees" reduced towel consumption 15%.

What's the takeaway? People respond to incentives. In this case, by using two paper towels instead of three, they felt like they were doing their job to help the environment. Non-monetary incentive, for sure, but an incentive nonetheless.



That link should have been http://thesecomefromtrees.blogspot.com

I always make that typo. Argh.


A rose by any other name...
How was the chocolate coffee...??


I went to New England Culinary Institute and they actually have a class on menu writing. It is truly an art when done correctly. The whole point is to discover the language that entices people, and makes them spend more money. Certain words (just like in real estate adds) have positive connotations. There is a grouping of words to signify value without making it sound cheap and a grouping to instill luxury and class. Language is a powerful tool. Visual appeal, of course, plays a large part as well. We eat with our eyes first - starting with the menu. Visual appeal can make a world of a difference!


pete, the takeaway from that certainly does not appear to be "people respond to incentives," unless the word incentives is construed so broadly as to be virtually meaningless (at least from the standpoint of empirical prediction).

your effect seems to be much more of a psychology thing. there are no real "incentives" here except a "warm glow" - neither money nor social sanction is involved, since presumably there's no observability of this particular act.


An unintended experiment - from the Modern Food cookbook of the Rebar restaurant in Victoria, BC:

"Relish - a relish or a salsa? Years spent writing out menus and describing food in the written form has taught us some valuable lessons about the power of language in sales. A good example is the use of the term "relish" as a menu descriptive. For years, this "sweet corn salsa" has been served alongside enchaladas, a staple menu item. During gone of our biannual menu revamp sessions, we decided to rename this side dish to "sweet corn relish". Suddenly, our popular enchiladas were not selling! When the connection was finally unearthed, we promptly reverted to the original "salsa". Sure enough, sales resumed."


I think menu experiments are a grande idea, which will result in venti knowledge [reference to Starbucks odd names for small/medium/large intentional]


My kid yesterday watched a show on PBS called Fetch with Ruff Ruffman, and the kids on the show made what they called BLT Ice Cream (with butternut squash, lemon and tomato). After perfecting the recipe, mostly by lowering the tomato content, they went out into the world (somewhere on the east coast) and set up a free sample stand asking the participators to rate their experience. After the first group, they renamed the milkshake (and changed the signage) to something like Summer Citrus Cooler and their average rating, attributable only to the name change, went from about 3.5 to 4.2 on a scale of 1 to 5.