same time period. Another type of market study relies on time-series data.
Here, the firm chooses a single geographic area and varies its key decision vari-
ables over time to gauge market response. The firm might begin by setting a
high price and a low advertising expenditure and observing the market
response. Some time later, it may increase advertising; later still, it may lower
price; and so on. Time-series experiments have the advantage that they test a
single (and, one would hope, representative) population, thus avoiding some
of the problems of uncontrolled factors encountered in cross-sectional stud-
ies. Whatever the type, traditional market tests and studies are expensive—
often extremely so. A very rough rule of thumb holds that it costs $1 million
and more to conduct a market test in 1 percent of the United States.
The last decade has seen an exponential increase in management’s use of
Internet-based controlled market tests. For instance, Google might set out to
test the effect of different Internet ads on customer click-through rates. By ran-
domly assigning Internet visitors to different types of ads, it can compare the
average response rate of 20,000 visitors seeing one ad to 20,000 visitors seeing
the alternative ad. Because the populations of randomly selected visitors will be
essentially identical, any difference in response rate can be attributed to the dif-
ferent ad treatment. The company Omniture provides clients immediate feed-
back on different Web-page designs by randomly testing specified alternatives
and immediately comparing their effectiveness. The credit-card company
Capital One does much the same thing when it runs controlled tests of differ-
ent credit-card solicitations, systematically varying features such as the interest
rate and cash-back percentage. Running and analyzing these newest kinds of
controlled experiments can be accomplished at ever decreasing costs.^3132 Chapter 4 Estimating and Forecasting DemandCHECK
STATION 1An airline is considering expanding its business-class seating (offering more room
and amenities to business travelers at a slightly higher fare than coach). Which
method, survey or controlled market study, would you recommend to gather infor-
mation for this decision?Uncontrolled Market Data
In its everyday operation, the market itself produces a large amount of data.
Many firms operate in multiple markets. Population, income, product features,
product quality, prices, and advertising vary across markets and over time. All
of this change creates both opportunity and difficulty for the market
researcher. Change allows researchers to see how changing factors affect(^3) For a detailed discussion of controlled randomized market tests, see I. Ayres, Super Crunchers: Why
Thinking-by-Numbers Is the New Way to Be Smart, Chapter 2 (New York: Bantam Dell Publishing Group,
2007). Of course, controlled market tests are not new. Fifty years ago, David Ogilvy, the lion of
advertising, extolled the virtue of mail-order advertising because its impact was immediately
testable. Either readers clipped the coupon, or they didn’t.
c04EstimatingandForecastingDemand.qxd 9/5/11 5:49 PM Page 132