Starbucks decided to raise its drink prices by as much as 8% (5 cents to 30 cents), They are doing this just when  customers are cutting back on their Starbucks trips and switching to cheaper alternatives from McDonalds and Dunkin Donuts.

The conventional “wisdom”  on pricing is, when recession pushes customers to cut  back on expenses and switch from your products to cheaper alternatives,  you cut your prices to keep the customers. While this is a usually accepted and followed practice, it is neither wisdom nor based on analysis.
To be successful, businesses cannot make decisions based on hunch, gut feel, latest management fad, or so called conventional wisdom. Decisions need to be based on data and analysis which is easier said than done.
In the case of Starbucks, how did they arrive at price increase, going against the flow? The simplest calculation here is, when price conscious customers moved out all they are left with are price insensitive customers who prefer their products. Hence it makes sense to charge more for them as long as the loss in profit from further drop in customers is less than the increase in profit from higher price. (Here is an attempt at formal proof on why increasing prices yields better profits).
Starbucks has a gross margin of 22%. This is however the average. On their high priced premium drinks we can assume that their margins are at least twice as much. So let us say it is 44% gross margin. Their premium drinks retail for $3.75 or higher, so the new price is $4.05 and at 44% margin, their profit per cup is$1.78 . (Please note that gross margin numbers from GAAP income statements are not based on just marginal costs and include fixed cost allocations.)
Let us say they sell ‘N’ premium drinks in a year at the current price. The increase in profit from 30 cent price increase (if the number of drinks sold remains ‘N’) is  0.3N. You will see the value of N is not important to the analysis.
However, there is bound to be fall in sales. But how far should the sales fall to negate the benefits of price increase?
Let us say the sales falls from  by ΔN cups, then lost profit from this lost sales is   1.78ΔN
Their price increase will result in net loss only if 1.78ΔN  > 0.3N, that is sales has to fall by 17% from its current levels.  One in six people has to stop buying the premium drink.  (Note: We assumed a 44% margin, if it is lower  that that then the sales have to drop much more than 17% to make the price increase option unattractive)
Another way to look at this is from price elasticity of demand – % change in volume for one % change in price. For the price increase to be unprofitable, price elasticity of demand must be just over 2 (every % increase in price should result in  drop in volume of 2%).

The New York Times asks, “Will the hard-core customers pay more”? How likely is a 17% sales drop? Not very given that most price sensitive customers have moved out and what they are left with are those who prefer Starbucks over other brands. So their price sensitivity is most likely to be lower than it would have been before the recession.
Hence a 17% drop in sales is  highly unlikely. In terms of elasticity, premium drinks moved to inelastic part of the demand curve. As long as the sales drop stays below the 17%  mark, the price increase is a profitable and a smart move for Starbucks.
You can see how  Starbucks would have made this counter-intuitive decision – because it is based on evidence and analysis. Unfortunately this does not come naturally to most marketers, because no one wants to go against the flow or stand-up to authority. Worse, most marketers accept conventional wisdom without a challenge  because they lack  inclination and wherewithal to seek the right data and do the relevant analysis.
How was your last marketing decision made?