About 15 years ago, taking a usability engineering class from Randy Pausch at UVa (yes, that Randy Pausch), I worked with a team that was determined to develop a "cheat sheet" to help CS students debug the awful HP Laserjet IV printers in the UVa CS lab.

Think "PC LOAD LETTER?! What the f* does that mean?!"

After several user rounds of user-testing, we found ourselves using similar tricks. While iterating on documentation that would maximize our users' success, we occasionally found it necessary to provide partial information or even outright lie to the user! While this sounds under-handed, it was done to (1) help the user accomplish their intended goal (e.g., fix the damn printer) and (2) ensure that the user read enough of the instructions to be able to perform the task.

In all seriousness, we found that most CS students, unsurprisingly, do not have the patience to RTFM. But we were ultimately successful in tricking them into doing the right things.

How we phrased and partitioned the branching instructions turned out to be more important than providing a deep understanding of the printer itself.

Then again, some of the solutions to HP LJ IV problems, derived from their awful awful manual, seem like voodoo. ;-)

The Ubuntu documentation could use some help.

The pasta is a concession to the value consumer, and while it's high margin, given high fixed costs the restaurant would likely go out of business if it sold nothing by inexpensive pasta dishes. Ingredient costs are around 25% of the menu price, higher for expensive places and lower for chains. Differences in prices between dishes are generally not equal to differences in the cost of ingredients.

From your example, your implication is that all dishes in that restaurant are overpriced. While this may certainly be true relative to the cost of making the food yourself (presuming free time and available equipment) this belief should be tempered by the number of profit-less restaurants that go out of business each year. There's nothing wrong with your belief, but I don't think it's that relevant to discussions of 'decoy pricing'.

This is an extreme example, but what I'm talking about is more common. Think about chain restaurants like CPK, Macaroni Grill, PF Changs, etc. These restaurants just need to keep you seated long enough to sell you drinks. In order to do that without losing money on dishes, they need to funnel you into higher margin entrees. All other things being equal (i.e. the fixed costs of keeping the restaurant open), they want you eating pasta, not chicken.

It's simple math...pasta costs pennies per serving in raw materials, can be stored for long periods of time, and is easy to prepare (you boil and simmer it, and it is easy to keep warm while you prepare other dishes). It is by far the highest margin food item on the menu. Salad would be close, but it has a short shelf life. Chicken and other meat dishes are more complicated because they have a short shelf life (you don't freeze these because it kills the flavor and texture), and they need to be cooked just right and served quickly or you risk serving it cold or drying it out under heat lamps (resulting in expensive do-overs). The $15 (or $20) chicken entree is the real decoy...the restaurant would love to serve everyone pasta because it is a much higher margin (all other fixed costs being the same).

EDIT: I just googled "restaurant profit margins" and the second result gives the same example: http://articles.moneycentral.msn.com/SavingandDebt/ConsumerA...

Given the slim profit margin, many restaurants rely on savvy pricing to create the illusion of value. Putting a chicken dish on the menu for $21 will make a $15 pasta dish, where the restaurant is making a big profit, seem like a bargain, says Gregg Rapp, the owner of consulting firm MenuTechnologies.net.

That's funny...good thing I'm not the only one who noticed that at restaurants!

> In Decoy Marketing, we learned that adding an inferior but similarly priced product

> to one’s lineup can boost sales of the better product, and in More Decoys: Compromise

> Marketing we found that an expensive product added at the top of one’s line may not

> sell well but can boost sales of the cheaper products.

Seems like you have to A/B test over time and not offer different prices simultaneously or else you risk pissing people off.

You could also play around with sales tax where one menu has values of say $7.99 (tax not included) and another has $9 (tax included)

Are there any solid statistics-backed intro books to Behavioral Economics?

[1] http://www.amazon.com/Predictably-Irrational-Revised-Expande...