Pricing Information Carl Shapiro Hal R. Varian Information Rules

Britannica v. Encarta  Britannica: 200 years, $1,600 for set  1992: Microsoft purchased Funk & Wagnalls to make Encarta  Britannica response Online subscription at $2,000 per year Sales dropped 50% between 1990 and 1996 Online subscription at $120 CD for $200, since 1996 $70-$125

Production Costs  First-copy costs dominate Sunk costs - not recoverable  Variable costs small; no capacity constraints Microsoft has 92% profit margins  Significant economies of scale Marginal cost less than average cost Declining average cost

Implications for Market Structure  Cannot be "perfectly competitive"  2 sustainable structures Dominant firm/monopoly Differentiated product  …and combinations of above

Strategy  What to do Differentiate your product  Add value to the raw information to distinguish yourself from the competition Achieve cost leadership through economies of scale and scope  Limited scope for cost reduction  Need to increase volumes

Commoditized Information  CD ROM phonebooks  1986: Nynex charged $10,000 per disk for NY directory  ProCD and Digital Directory Assistance  Chinese workers at $3.50 daily wage  Bertrand competition Start at $200 each Price forced to marginal cost

If You are in Commodity Business  Cost leadership  Sell the same thing over again Reuters Reduces average cost

Differentiate Product  Bigbook and maps  West Publishing and page numbers

First-mover Advantages  Avoid greed Respond to threat quickly and decisively pricing; highly credible with high FCs  Play tough Discourage future entry Constant innovation (search engines)

Hard to do for Incumbent  May not recognize threat till too late CP/M Wordstar VisiCalc

Personalize Your Product  Personalize product, personalize price Personalized ads

Know Your Customer  Registration Required: NY Times Billing: Wall Street Journal  Know your consumer Observe Queries Observe Clickstream

Forms of Differential Pricing  Personalized pricing Sell to each user at a different price Customers may not like it !  Versioning Offer a product line and let users choose  Group pricing Based on group membership/identity

Personalized Pricing  Catalog inserts Market research Differentiation  Easy on the Internet

Traditional Industries  Airlines  Supermarket scanners Profit margin more than doubled More effective than other forms of advertising

Internet  Auctions  Closeouts, promotions

Group Pricing  Price sensitivity  Network effects, standardization  Lock-In

Price Sensitivity  International pricing US edition textbook: $70 Indian edition textbook: $5  Problems raised by Internet

Network Effects  Compatibility Site licenses Variety of schemes: per client, per user, per server, etc.  Lock-In  Microsoft Office Per seat, concurrent

Sharing  Transactions cost of sharing  Desire for repeat play

Summary  Understand cost structure  Commodity market: be aggressive, not greedy  Differentiate product and price  Understand consumer  Personalize products and prices  Consider selling to groups

Versioning Information Carl Shapiro Hal R. Varian

Value-Based Pricing  Don’t need to price by identity  Offer product line, and watch choices  Design menu of different versions Target different market segments Price accordingly (self selection)

Quicken Example Revisited  Quicken for Windows at $20  Quicken Deluxe at $60

Traditional Information Goods  Hardback/paperback  Movie/video

Dimensions to Use  Delay (Fed Ex, PAWWS)  User Interface (DialogWeb, DataStar)  Image Resolution (PhotoDisk)  Speed of operation (Mathematica)  Format (Lexis/Nexis)  Capability (Kurzweil)  Features (Quicken, tech support)  Comprehensiveness (DialogWeb, DataStar)  Annoyance  Support

Example  40 type As: $100 for speed, $40 for slow  60 type Bs: $50 for speed, $30 for slow  Identity-based pricing: $7000 revenues  Offer only speedy: $50 is best price, revenues=$5,000  Offer only slow: not as profitable

Versioning Solution  Try speedy for $100, slow for $30 Will this work? Compare benefits and costs =0, but 40-30=10 > 0 Discount the fast version: 100-p=40-30 So, p=90 Revenues = $5,400 = 90x x60

Making Self-Selection Work  May need to cut price of high end  May need to cut quality at low end  Value-subtracted versions May cost more to produce the low- quality version.  In design, make sure you can turn features off!

Pitfalls  Arbitrage/ Turning low-end version into high-end version Windows NT workstation  Can run server with limited number of concurrent users server

Online and Offline Versions  Does the on-line version stimulate demand or steal from the offline version  Economist, WSJ Free online version for print version subscribers (complement) Free print version for on-line subscribers (substitute) Marginal costs? Vs. benefits

How Many Versions?  One is too few  Ten is (probably) too many  Two things to do Analyze market Analyze product

Analyze Your Market  Does it naturally subdivide into different categories? AND  Are their behaviors sufficiently different?  Example: Airlines Tourists v. Business travelers

Analyze Your Product  Dimensions to version  High and low end for each dimension  Design for high end, reduce quality for low end  Low end advertises for high end

Goldilocks Pricing  Mass market software (word, spreadsheets) Network effects User confusion  Default choice: 3 versions  Extremeness aversion  Small/large v. small/large/jumbo

Microwave Oven Example  Bargain basement at $109, midrange at $179 Midrange chosen 45% of time  High-end at $199 added Mid-range chosen 60% of time  Wines Second-lowest price

Customizing the Browser  Collect behavior information (Client- side code)  Optimize viewing B&W page images Buffering  Can turn it on and off

Bundling  Offer a package  Microsoft Office 90% market share  Work together  Discount one of the products  Option value: zero incremental price  Microsoft's per-processor license  Use Bundling to introduce new products to consumers

Reduce Dispersion: Price separate or together? Profits: With Bundling: $440Without: $400

Information Bundles  Magazines and newspapers  Economies of scale  People are willing to pay for reading some articles some of the time  Different users want different bundles Noah & Mark – S1(1.20,1), S2(1,1.20)  Customized bundles  Nonlinear pricing Pricing music  $9.99 for x downloads for each additional  Reduced dispersion results in increased revenue

Promotional Pricing  Sales, coupons, rebates  Only worthwhile if segment market  Each target group needs to “exercise some effort” to use the coupon  Credible signal of price sensitivity  software agents Shopping.com Epinions.com

Lessons  Version your product  Delay, interface, resolution, speed, etc.  Add value to online information  Use natural segments  Otherwise use 3  Control the browser  Bundling may reduce dispersion

Rights Management Carl Shapiro Hal R. Varian

Intellectual Property Law  “Intellectual property law cannot be patched, retrofitted, or expanded to contain digitized expression…Information wants to be free.” John Perry Barlow  Is he right?

Production and Distribution  Digital tech lowers production costs  Digital tech lowers distribution costs  Examples Tape recorder lowers production, but not distribution costs AM radio broadcast lowers distribution costs, not reproduction costs  Current digital technologies lower both

Make Lower Distribution Costs Work for You  Information is an experience good  Must give away some of your content in order to sell rest  Can use product line/versioning National Academy of Sciences Press Easy to read, hard to print On-line versions boost sales of off-line versions

Demand for Repeat Views  Give away all your content, but only once  Generate demand for repeat views  Music, books, video have different use patterns  Children Barney: free videos Disney: sued day care centers  Adults Video rentals vs. video purchase

Demand for Similar Views  Free samples direct customers back to you  McAfee Associates $5 million in first year $3.2 billion market value by 1997 Half of virus protection market

Demand for Complementary Products  Give away index and sell content Wall Street Journal, New York Times, Economist give away index, Books on Amazon

Illicit Copying  Timely information: not a big problem  Cheap information: not a big problem  Negative feedback: the bigger you are, the easier to detect  Digital Watermarking  Digital Rights Management tml tml

Lower Reproduction Costs  Perfection isn’t as important as commonly thought  Digital Audio Tape (DAT) SCMS inhibits copies of copies  Analog video tapes: 1979: 4 blanks for each pre-recorded 1992: 1 to 1

Issues  Patent battles  Standards battles  Inconvenience Spreadsheet copy protection  Price of content  Reliability Technical and procedural

Historical Examples  Circulating libraries 1741: Pamela 1000 libraries by 1840  Video stores Video rental as prelude to purchase Growing the market

Transactions Costs  Site license v individual licenses? Who can distribute more cheaply? How effectively can group aggregate value?

Lessons  Two challenges: cheap production, cheap distribution  Cheap distribution: helps advertise by giving away samples  Cheap distribution: good for bitleggers, but their need to advertise helps control them

Lessons, continued  Copy protection that imposes costs on users is vulnerable to competitive forces  Basic tradeoff in terms and conditions: more liberal terms make product more valuable but may reduce sales  Site licenses and other group pricing schemes are a valuable tool

Case Study: Netflix  Bundling  Transaction Costs