Demand Analysis and Optimal Pricing 103purchase fewer units at a higher unit price. Perhaps the most common form of
quantity discounts is the practice of two-part pricing.As the term suggests, the
total price paid by a customer iswhere A is a fixed fee (paid irrespective of quantity) and p is the additional
price per unit. Telephone service, electricity, and residential gas all carry two-
part prices. Taxi service, photocopy rental agreements, and amusement park
admissions are other examples. Notice that two-part pricing implies a quantity
discount; the average price per unit, P/Q A/Q p, declines as Q increases.
Two-part pricing allows the firm to charge customers for access to valuable serv-
ices (via A) while promoting volume purchases (via low p).Information Goods
In the last 20 years, we have witnessed explosive growth in the provision of
information goods and services. The business press speaks of Internet indus-
tries and e-business markets. The “information” label is meant to be both more
broad based and more precise. An information good could be a database, game
cartridge, news article (in electronic or paper form), piece of music, or piece
of software. Information services range from e-mail and instant messaging to
electronic exchanges and auctions, to brokerage and other financial services,
to job placements. Of course, information services also include all manner of
Internet-based transactions, such as purchasing airline tickets, selling real
estate, procuring industrial inputs, and gathering extensive data on potential
customers.^17
Although the information category is broad, all of the preceding exam-
ples share a common feature:Information is costly to produce but cheap (often cost-
less) to reproduce. In short, any information good or service is characterized by
high fixed costs but low or negligible marginal costs. With marginal costs at or
near zero, the firm’s total costs vary little with output volume, so that average
cost per unit sharply declines as output increases. (Creating a $1 million data-
base to serve 1,000 end-users implies an average cost of $1,000 per user. If,
instead, it served 500,000 end-users, the average cost drops to $2 per user.)
Moreover, with marginal costs negligible, a supplier of an information good
once again faces a pure selling problem: how to market, promote, and price its
product to maximize revenue (and thereby profit).
Not surprisingly, the early history of e-business activities has been charac-
terized by high up-front costs and the pursuit of customers, revenues, and prof-PApQ ,(^17) A superb discussion of the economics of information goods can be found in C. Shapiro and
H. R. Varian, Information Rules, Chapters 1–3, 7 (Boston: Harvard Business School Press, 1999).
c03DemandAnalysisAndOptimalPricing.qxd 8/18/11 6:48 PM Page 103