ONLINE AUCTIONS AND MULTICHANNEL SALES PROCESSES: THE ROLE OF SELLER - - PDF document

• 1 -

ONLINE AUCTIONS AND MULTICHANNEL SALES PROCESSES: THE ROLE OF SELLER CAPABILITIES Jason Kuruzovich The Lally School of Management and Technology Rensselaer Polytechnic Institute, Troy, New York 12180 518.276.2332; kuruzj@rpi.edu

• 2 -

PAPER SUMMARY A key advantage of online auctions over traditional live auctions is the decoupling of the information and logistics components of the transaction (Van Heck and Ribbers 1997). This decoupling reduces the transportation costs associated with offering an item to a separate market while enabling sellers to simultaneously list products in online auctions and other channels. The use of online auctions along with other channels is increasingly prevalent for items which are unique and have a high valuation in comparison to the transaction costs involved—i.e., private jets, boats, vehicles, real estate, bicycles, etc. While a great deal of research has examined the relevant drivers of price in online auctions (Bajari and Hortacsu 2002; Bapna et al. 2001; Lucking-Reiley et al. 2007), there is a need to further understand how sellers actually utilize

• nline auctions as part of the retailing operation (Pinker et al. 2003).

This research asks the question, how are online auction channel outcomes influenced by seller resources and capabilities to sell products across alternate channels? Our approach involves using data collected from eBay Motors to characterize channel outcomes for a set of sellers, showing how heterogeneity in seller capabilities drives channel outcomes. A high percentage of sellers in this market are professional car dealers who are differentiated in their ability to locate buyers outside of the online auction channel. In particular, we characterize differences in the electronic commerce capabilities of sellers, and we find that seller capabilities are directly related to the relative prices the sellers obtain (controlling for item characteristics), the percentage of vehicles which end in a sale (closing ratio), and the percentage of vehicles listed which are sold through the online auction channel (sell-through rate). An overview of this model is shown in Figure 1. Jointly examining these three outcomes provides insights into how sellers’ utilize online auctions as part of a multichannel sales process.

• 3 -

A key contribution of this paper is to empirically show that online auction channel

• utcomes are influenced by seller’s capabilities related to the retail channel, thus presenting a

multichannel view of the use of online auctions and highlighting the role of seller heterogeneity in driving online auction outcomes. The current manuscript is in initial draft form, with initial findings as noted. I am currently adding geo-coded data incorporating the degree that the location of the sellers acts as a resource that drives seller outcomes and will have results and a fully developed working paper by the conference. METHODOLOGY Data were collected from eBay motors over a period of six months. Data accessing retailer electronic commerce capabilities were subsequently collected from retailer websites. In controlling for vehicle characteristics, we limited our analysis to only vehicles which were sold in the year 2000 or later and models (determined through the first 8 digits of the VIN) with at least 30 sales. We estimated the relative revenue outcome of sellers (AVGPj) by averaging the difference between that actual end price of the auctions for vehicles that were sold (Pi) and the expected sale price (pi) controlling for vehicle characteristics (vi) and the timing of the auction (ti) for each seller (j): We limited our analysis to sellers who had sold at least 5 vehicles. (3) The closing ratio (CR) for the each seller was estimated by dividing the total number of unique vehicles sold through the auction channel (nj(SOLD)) by the total number of auctions (nj(AUCTIONS)) the seller participated in.

• 4 -

(4) The sell through rate (STRj) was calculated by the dividing the total number of unique vehicles sold through the auction channel (nj(SOLD)) by the total number of unique vehicles listed (nj(LISTED)). (5) Electronic commerce capabilities were measured using data collected from the websites

• f sellers. A listing of 8 relevant website characteristics (wc) of interest to auto retailing was

adapted from a survey created by J.D. Power and Associates, a leading market research firm within the auto industry. Research assistants located electronic commerce websites and noted the whether the retailer had incorporated functionality such as “Lists Price on Website,” and “Schedule Vehicle appointment.” This mechanism of capturing specific IT functionality relevant for an industry is consistent with prior work measuring electronic commerce capabilities of retail

• rganizations (Zhu and Kraemer 2002). For each seller the website characteristics were summed

according to the following equation: (6) We further characterize sellers electronic commerce capabilities through a Google search

• f each 17 digit VIN offered by the seller. This Google search returned a number of filtered

pages returned. By removing websites associated with eBay, this measure gives us a way to characterize other websites in which the same vehicle has been listed, or the sellers Search in Other Websites (SOW) for each auction. This measured the sellers use of other websites, both

• 5 -

third parties and the dealers own website, to market vehicles. For each seller, we calculated the average search in other websites (AVGSOWj) across all auctions and vehicles listed. (7) We further controlled for the reputation of the seller. Following prior work in this area we used the number of positive comments (PR) and the number of negative comments (NR) separately. The full system of equations in relating the characteristics of the seller to auction related

• utcomes is found below.

(8) (9) (10) To obtain efficient estimators of the parameters of the equations the analysis used seemingly unrelated regression (SUR). RESULTS Descriptive statistics and correlations between seller characteristics are found in Table 1. The regression model controlling for the predicted sale price included 13,160 auctions, and the results are found in Table 2. SUR estimates for equations 8-10 for 322 sellers are found in Table

• 3. Overall results indicated a relationship between electronic commerce capabilities and the

average price, the closing ratio, and the sell through rate, as suggested by search theory. Results also show that reputation (as measured by feedback) is positively associated with the average price, the closing ratio, and the sell through rate. Overall, the findings support the important role

• f seller capabilities in influencing online auction outcomes.

• 6 -

Table 1. Descriptive Statistics and Correlations Mean Std 1 2 3 4 5 6

1

ECC 3.69 2.94

2

AVGSOW 1.43 0.84 0.27

3

PR 258.02 432.3

• 0.10
• 0.04

4

NR 2.58 4.98

• 0.18
• 0.13

0.58

5

AVGP

• 0.04

1.85 0.21 0.26

• 0.03
• 0.10

6

STR 0.51 0.30

• 0.36
• 0.46

0.06 0.10

• 0.34

7

CR 0.35 0.26

• 0.31
• 0.44

0.04 0.11

• 0.36

0.93

• 7 -

Table 2. Regression Analysis of Sale Price Variable Sale Price1 ln(Miles) (miles) -1.174***(0.038) Certified (0/1)

• 0.102***(0.262)

Inspected (0/1) 0.037***(0.068) Warranty (0/1) 1.563***(0.088) Time (weeks)

• 0.018***(0.006)

R2 0.931 N 13160

Note: * p<0.05 **P<0.01 ***p<0.001

1Dummy variables for vehicle model (425), vehicle year (7), and color (6) are not shown.

Table 3. SUR Analysis Variable Average Price (AVGP) Closing Ratio (CR) Sell Through Rate (STR) ln(ECC) 0.363**(0.120)

• 0.038**(0.010)
• 0.065***(0.014)

ln(AVGSOW) 0.526*(0.208)

• 0.130***(0.018)
• 0.163***(0.024)

ln(PR) 0.099 (0.101) 0.034***(0.009) 0.069***(0.012) ln(NR)

• 0.429**(0.137)
• 0.007(0.012)
• 0.000576

R2(OLS) 0.113 0.268 0.297 R2(SUR) 0.16 N 322

Note: * p<0.05 **P<0.01 ***p<0.001

• 8 -

References Bajari, P., A. Hortacsu. 2002. The winner's curse, reserve prices, and endogenous entry: Empirical insights from eBay auctions. The Rand Journal of Economics 34(2) 329-356. Bapna, R., P. Goes, A. Gupta. 2001. Insights and analysis from online auctions. Communications

• f the ACM 44(11) 42-50.

Genesove, D. 1995. Search at wholesale auto auctions. The Quarterly Journal of Economics 110(1) 23-49. Lucking-Reiley, D., D. Bryan, N. Prasad, D. Reeves. 2007. Pennies from eBay: The determinants of price in online auctions. Journal of Industrial Economics 55(2) 223-233. Pinker, E. J., A. Seidmann, Y. Vakrat. 2003. Managing online auctions: Current business and research issues. Management Science (49:11) 1457-1484. Van Heck, E., P. M. Ribbers. 1997. Experiences with electronic auctions in the Dutch flower

• industry. Electronic Markets 7(4) 29-34.

Zhu, K., K. L. Kraemer. 2002. E-commerce metrics for net-enhanced organizations: Assessing the value of e-commerce to firm performance in the manufacturing sector. Information Systems Research 13(3) 275-295.

• 9 -

Figure 1. Research Model

• Avg. Sale Price

Sell Through Rate Closing Ratio Seller Resources and Capabilities R E S E R V E