1. 1 Teck-Hua Ho April 25, 2006 Pricing and Consumer Search I. Economic and Behavioral Foundations of Pricing II. Innovative Pricing Concepts and Tools III. Internet Pricing Models

2. 2 Teck-Hua Ho April 25, 2006 Outline  Convergence Hypothesis  Price Level: Bertrand Paradox  Price Dispersion: Economics of Search  Search Agents  Field Evidence on Price Levels and Dispersion  A Taxonomy of Competition

3. 3 Teck-Hua Ho April 25, 2006 Convergence Hypothesis “The explosive growth of the Internet promises a new age of perfectly competitive markets. With perfect information about prices and products at their fingertips, consumers can quickly and easily find the best deals. In this brave new world, retailers’ profit margins will be competed away, as they are all forced to price at cost.” The Economist, November 20, 1999, p. 112.

4. 4 Teck-Hua Ho April 25, 2006 Two Predictions about Internet Price Level and Price Dispersion  Prediction 1:  Price = Marginal Cost  Retailers make zero profit  Prediction 2:  All retailers charge the same price  Zero price dispersion

5. 5 Teck-Hua Ho April 25, 2006 Price Level: Bertrand Paradox  A group of homogeneous consumers  There are two retailers who sell an identical good (e.g., “Strategic Pricing” book)  The two retailers have an identical marginal cost and are located next to each other. Consequently, consumers buy from the retailer who charges the lowest price  Bertrand shows that each retailer has a strong incentive to undercut the price of the other retailer (to capture the whole market) until both prices hit the common marginal cost  Price = Marginal Cost  Retailers make zero profit (even when there are only 2 firms competing)

6. 6 Teck-Hua Ho April 25, 2006 Price Dispersion: Economics of Search  Price of Olympus D-360L Digital Camera:  Assume 50% of the stores price it at $300 and 50% of the stores price it at $200  The costs of search per store visit is $X  If the shopper visits only one store, she has a 50% chance of buying the product at $200 and a 50% chance of buying the product at $300. So the expected price is $250.  What is the expected price if she visits 2 stores?

7. 7 Teck-Hua Ho April 25, 2006 Expected Price versus Number of Searches  She will have a 75% chance of buying the product at $200 and a 25% chance of buying it at $300. So the expected price is $225. In general, we have: Number of Searches Probability buying at $200 Probability buying at $300 Expected Price 10.5 $250 20.750.25$225 30.8750.125$212.5 40.93750.0625$206.25 50.968750.03125$203.125

8. 8 Teck-Hua Ho April 25, 2006 Optimal Number of Search: The Impact of Search Cost  If $X is $10, what is the optimal number of search?  If $X is $5, what is the optimal number of search?  If $X is $0.25, what is the optimal number of search? Number of Searches Probability buying at $200 Probability buying at $300 Expected Price 10.5 $250 20.750.25$225 30.8750.125$212.5 40.93750.0625$206.25 50.968750.03125$203.125 3 4 8

9. 9 Teck-Hua Ho April 25, 2006 The Role of Search Agents  Encourages cross-store comparison and lowers the search costs for price information  Increases consumer price sensitivity

10. 10 Teck-Hua Ho April 25, 2006 A Comparison of Search Agents  www.MySimon.com www.MySimon.com  www.DealTime.Com www.DealTime.Com  www.Bizrate.com www.Bizrate.com ProductMySimonDealTimeBizRate Canon PowerShot S2 IS (Best) $329.00 $312.00 Canon PowerShot S2 IS (Worst) $447.00 $399.99  Further reduces the search costs for price information!

11. 11 Teck-Hua Ho April 25, 2006 Summary  Prediction 1:  On internet, retailers who are selling identical goods and who are “next to each other” (in the e-space), engage in Bertrand price competition. Consequently, prices will be driven down to marginal cost.  Prediction 2:  On internet, search cost  $0. Customers will all buy from the the cheapest stores and the more expensive stores will be driven out of the market.

12. 12 Teck-Hua Ho April 25, 2006 Field Evidence on Price Levels and Dispersion  Field study of books and CDs *  Price levels: 9-16% lower on Internet than in B&M  Dispersion of posted prices: 25-33%, larger than in B&M (books only)  Dispersion of “share-weighted” posted prices: lower than in B&M  Cheapest e-tailers did not get largest market share  Field study of online travel agents **  The same customer request led to tickets offered with substantially different prices and characteristics, 25% dispersion  Even after controlling for differences in ticket characteristics, 20% dispersion * Erik Brynjolfsson and Michael D. Smith, “Frictionless commerce? A comparison of Internet and conventional retailers,” Mgmt Sci, April 2000; ** Eric K. Clemons, Il-Horn Hann and Lorin M. Hitt, “The nature of competition in electronic markets: An empirical investigation of online travel agent offerings,” Working paper, The Wharton School, 1998.

13. 13 Teck-Hua Ho April 25, 2006 Two Challenges to the Predictions  Non-price competition: Consumers care about other product attributes besides price. That is, products are always differentiated to some degree (e.g., brand, can be delivered sooner, better return policies, comes with superior post-sale services)  Search costs for quality information: If search costs for quality information are also lowered, customer price sensitivity may decrease. The net effect depends on the relative size of reductions in search costs for price and quality information

14. 14 Teck-Hua Ho April 25, 2006 Search Costs versus Nature of Competition Search Costs for Quality Information Search Costs for Price Information LowHigh Low High Price Competition Quality Competition Value Competition Minimal Competition

15. 15 Teck-Hua Ho April 25, 2006 High Search Costs for both Price and Quality Information  B&M environment (e.g., buying a T-shirt on the streets of Bangkok, Thailand)

16. 16 Teck-Hua Ho April 25, 2006 Low Search Costs for both Price and Quality Information www.peapod.com 06001

17. 17 Teck-Hua Ho April 25, 2006 Field Evidence on Price Elasticity at Peapod * Attributes for which more information is accessible online are more salient Sensory search attributes for which less information is available online are less salient Brand names are more important online, at least for products for which little attribute information is available online The combined effect of price and promotion is weaker online * Alexandru Degeratu, Arvind Rangaswamy and Jianan Wu, “Consumer choice behavior in online and traditional supermarkets: The effects of brand name, price and other search attributes,” Int J of Res in Mktg, March 2000.

18. 18 Teck-Hua Ho April 25, 2006 Wine Online: Experimental Design There are 100 products (indexed by 1,2,3,…, 98, 99, 100) and 2 online stores Store 1 carries 1, 2, …, 39, 40, 81, 82,.., 99, 100 and store 2 carries 41,42,.., 79, 80, 81, 82, …, 99, 100 Three independent variables; 1) Price-Usability (High or Low) 2) Quality-Usability (High or Low) and 3) Store Comparability (High or low) resulting in a 2x2x2 factorial design 72 participants, each is assigned to one of the 8 treatment conditions and makes 8 shopping trips Each product is on deal (15% lower) on four of the shopping trips

19. 19 Teck-Hua Ho April 25, 2006 Wine Online: Dependent Variables  Price sensitivity  Price elasticity  Quantity difference  Rated liking of the purchased wines  Retention: Willingness to subscribe to the same electronic-shopping wine service from homes  Market share of the common brands

20. 20 Teck-Hua Ho April 25, 2006 Customer Price Sensitivity : Quality Usability

21. 21 Teck-Hua Ho April 25, 2006 Customer Price Sensitivity: Store Comparability

22. 22 Teck-Hua Ho April 25, 2006 Rated Liking of Purchased Wines

23. 23 Teck-Hua Ho April 25, 2006 Retention  Quality-Usability  Low (29%) and High (49%)  Store Comparability  Low (27%) and High (50%)

24. 24 Teck-Hua Ho April 25, 2006 Market Share of Common Brands: Distribution  Common brands (Products 81, 82, …, 99, 100)  Store Comparability  Low (35.9%) and High (26.2%)  Benchmark 1: 33.3%  26.2 % is statistically smaller than 33.3%  35.9% is not statistically different from 33.3%  Benchmark 2: 20%  26.2% is statistically higher than 20.0%  What do these results mean for manufacturers?

25. 25 Teck-Hua Ho April 25, 2006 Punchline Search Costs for Quality Information Search Costs for Price Information LowHigh Low High Price Competition Quality Competition Value Competition Minimal Competition