The Role of Longevity


As in other markets, a seller's longevity in the market provides a positive signal regarding the seller's trustworthiness . Trustworthy sellers often provide signals to buyers regarding the quality of the good or service they provide. Brick and mortar investments, for example, provide incentives for companies to be reputable in a way that is observable by potential customers. Online sellers are generally unable to invest in their reputations with brick-and-mortar investments (with some 'brick-and-click' exceptions), but sellers can invest in their reputation via their record in past transactions. As Resnick, Zeckhauser, Swanson, and Lockwood (2003) point out, a good reputation is costly to maintain and so should allow the seller to do better than otherwise . Increasing seller experience in the marketplace , by increasing the seller's investment in his or her reputation, should enhance a buyer's expectations regarding the quality of the seller's goods and services. Cabral and Hortacsu (2002), for example, find that a combination of the number of transactions, age of membership, and eBay feedback rating has a 'more robust relation- ship with price.'

Research Questions and Methodology

In May and June of 2003, we collected nearly 300 observations from completed eBay auctions. As we discuss previously, the probability r that a seller provides a current buyer with a high quality product and good service is most likely equal to the probability that previous buyers had positive experiences with the seller. Thus, we regress price on the proportion of responses that are positive. To our knowledge, this has been previously done only in Kalyanam and McIntyre (2001), which uses the proportion of negative responses.

We collected data for two products: DVDs and designer watches. They were selected because they represent two different levels of uncertainty for buyers. DVDs are standardized, low risk, and low-cost items while designer watches are expensive (all winning bids exceeded $100) 'high touch' items.

As described in Table 7-2, variables collected from eBay listings included winning bid price, shipping cost, seller's reputation (rating), and seller experience level (estimated by number of past ratings). We also examined the listings to ensure that items satisfied this study's criteria that items must be new or in 'like new' or ' excellent ' condition as defined in Table 7-2. Finally, we include only items from completed auctions. Auctions with no bids or those in which the highest bid price did not meet the seller's minimum 'reserve' price were excluded from the data set.

Table 7-2: Description of variables

eBay Price: Price of winning bid for DVD/Watch. Includes only completed auctions where bid price exceeds 'reserve price' (the minimum price specified by the Seller).

Benchmark Price: Price of new DVD/Watch offered by Amazon.com or a Yahoo Shopping retailer.

Actual Shipping: Buyer's actual shipping cost.

Retail Shipping: Shipping cost if item had been purchased on Yahoo Shopping or Amazon.com. Note: Amazon.com charges $2.98 for DVD shipping while most Yahoo Shopping retailers offer free shipping for watches that exceed $100.

Total Price Ratio : (eBay Price + Actual Shipping) divided by (Retail Price + Retail Shipping)

Seller's Experience : Estimated by number of feedback ratings.

Seller's Reputation : eBay's Seller Rating (Number of Positive Ratings) divided by (Number of Positive + Number of Negative Ratings)

Description Quality: A dummy variable indicating the level of detail provided by the Seller in the item's description. High Ratings (or '1') were given to listings that met the following criteria:

* For watches: description included detailed images and text descriptions, model number, warranty information and formatted auction page.

* For DVDs: description included information about DVD's condition, DVD's details (format, movie description), image of cover, guarantees , and formatted page. * See Figures 7-4 and 7-5 in the appendix for examples.

We also include a measure of the seller's experience, proxied by the seller's number of previous feedback responses - generally the only measure of seller experience that buyers on eBay observe. This includes positive, neutral, and negative feedback responses. This is a truer measure of total experience than the number of positive responses alone. [2]

On an exploratory basis, we provide a rating of the item's description. All eBay sellers are given space to describe the item, post images, format the page, provide payment and shipping information and so on. In our view, this space serves as an important mechanism to promote the item and convey trust. As a result, we believe this is an aspect that warrants further consideration in eBay research. Thus, in an initial effort to explore the relationship between the quality of these descriptions and auction prices, we coded very detailed or well formatted descriptions as 'High Quality' descriptions. Details of the criteria used are described in Table 7-2 and examples for both watches and DVDs are found in the appendix of this paper.

Because prices of DVDs and designer watches vary by item, we accessed leading Internet retailers Amazon.com and Yahoo! Shopping to find benchmark prices. Amazon.com is among the leading DVD retailers while Yahoo! Shopping has relationships with many of the leading designer watch retailers. As a result, we computed a Total Price Ratio that represents the percentage a buyer paid for an item using eBay compared to the price the user would have paid at Amazon.com or a Yahoo! Shopping retailer. Only items for which the identical product listing could be found were included in the data set.

Statistical Analyses and Findings

Table 7-3 summarizes the descriptive data associated with the market for DVDs and designer watches on eBay. Note that eBay buyers of both items pay close to 70% of what they would pay for a new item had they purchased it from Amazon.com or a Yahoo Shopping retailer. We also see that designer watch sellers tend to have slightly lower seller ratings than DVD sellers. This probably stems from greater heterogeneity among designer watches than DVDs. Consistent with this, we find that a greater percentage of designer watch sellers create detailed product descriptions in their auction listings. Lastly, shipping charge data reveals that eBay sellers tend to charge more than their respective counterparts at amazon.com or Yahoo shopping retailers.

Table 7-3: Descriptive statistic

Variable

Designer Watches

DVDs

Mean Winning Price

$791.61

$9.92

Mean Benchmark Price 1

$1159.98

$17.35

Mean Shipping Charge 2

$12.82

$3.39

Mean Seller Rating (% positive)

98.47%

99.19%

Mean Number of Ratings

678.21

6981.38

Percent with 'high quality' descriptions

66.12%

37.21%

Mean Total Price Ratio 3

71.21%

67.80%

1: Based on new items offered by Amazon.com or a Yahoo Shopping vendor

2: Actual shipping charged as described in item listing. Amazon.com charged 2.98 for DVD shipping while Yahoo Shopping vendors typically offered free shipping for expensive designer watches

3: Computed by: (Winning Price + Actual Shipping) / (Benchmark Price + Benchmark Shipping)

As shown in Figure 7-4, items in which sellers presented 'high quality' descriptions (as coded by the authors) resulted in appreciably higher total price ratios for both DVDs (78% vs. 61%) [for examples of high quality descriptions, see the Appendix, Figures 7-5 and 7-6] and designer watches (76% vs. 61%). These results are somewhat surprising for the DVDs. This suggests it pays sellers to take advantage of eBay's description space to promote products. The results are exploratory and any conclusions are preliminary. There is a strong role for future research to examine with greater rigor the decision to promote and the impacts of promotion.

click to expand
Figure 7-4: Impact of description quality on winning bid prices

Tables 7-4 and 7-5 examine the correlations for DVDs and designer watches. Because they are heterogeneous items (i.e., different styles of watches and different DVD titles), we focus on the price ratio rather than the absolute value of the winning bid or benchmark price.

Table 7-4: Pearson correlation statistics for DVDs (n=172)
click to expand
Table 7-5: Pearson correlation statistics for designer watches (n=117)
click to expand

One important finding here is that the correlation between Seller Rating and Total Price Ratio is strong for designer watches and weak for DVDs. The Number of Seller Ratings is significantly correlated with Total Price Ratio for both items. Likewise, High Description Quality also correlates positively with Total Price Ratio for both types of items.

As one would expect, there is a strong positive correlation between Winning Bid Price and Benchmark Price. However, the relationship is much stronger for watches than for DVDs, indicative of the price variation for different DVD titles or genres. For example, a classic movie or kids ' title might earn a much higher percentage of its retail value on eBay compared to winning bid prices for other DVD titles. This would lead to another result we observe above: more expensive DVDs - but not more expensive watches - result in a lower price ratio.

As shown in Table 7-6, we regress Total Price Ratio on independent variables Seller Rating, Number of Ratings, and Description Quality. The F-statistics are highly significant for both items, with the combined variables explaining about 30% of the variation in designer watch prices and 17% of that in DVD prices. All three variables are significant. For designer watches, Seller Rating had a very high t-value (5.10) while Number of Ratings and Description Quality were also significant at (p<=.05). For DVDs, Description Quality had a high t-value of 3.77 while Seller Rating and Number of Ratings were only marginally significant (p<=.10).

Table 7-6: Multiple regression
click to expand

[2] In the context of our regression model, the impact of experience is estimated assuming the proportion of positive responses constant, so an increase in the number of negative feedback responses would come with a (generally substantial) increase in the number of positive responses. For example, if the proportion of positive responses were 95%, an increase in the number of negative responses by one would be accompanied by an increase of 19 in the number of positive feedback responses.




Contemporary Research in E-marketing (Vol. 1)
Agility and Discipline Made Easy: Practices from OpenUP and RUP
ISBN: B004V9MS42
EAN: 2147483647
Year: 2003
Pages: 164

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net