Daria Dzyabura, Srikanth Jagabathula and Eitan Muller (2017), "Accounting for Discrepancies between Online and Offline Product Evaluations."
October 2017. Read paper.
Most preference-elicitation methods that are used to design products and predict market shares, such as conjoint analysis, are conducted online. However, many firms sell their products offline, or through a mixed online-offline channel. In this paper, we demonstrate that large discrepancies can exist between attribute partworths when evaluating physical products versus online descriptions. We propose the following two-step solution for accurate estimation: (1) a data-collection method that combines an online study completed by a large number of respondents with an offline study completed by a small subset of the respondents, and (2) a statistical data-fusion method to estimate offline parameters by combining the online and the offline data. 

Eyal Biyalogorsky, Amir Heiman and Eitan Muller (2017), "Branding and the Ravages of Time: The Effect of Time on the Brand Premiums of Automobiles."
September 2017. Read paper.
We present a dynamic analytical model and empirical study of durable goods market with status consciousness of some consumers and show that as the importance of status increases, price of the used product decreases faster. We then use data on prices of new and used cars including the car’s age; its usage (distance driven); its external condition, and its status (premium vs. standard) for 21 twin car pairs, to estimate the deprecation in car values. The main result is that a premium car’s age depreciation is much higher than that of the standard car (controlling for their respective mileages and initial prices). This demonstrates that the true cost of owning a premium car is not just its initial high price, but the faster depreciation of the car’s intangible value over its lifetime.

Gil Appel, Barak Libai and Eitan Muller (2017), "On the Monetary Impact of Fashion Design Piracy."
July 2017. Read paper.  
We combine data collected on the growth of fashion items with industry statistics, to create a formal analysis and simulations of the monetary impact of a design pirated item (“knockoff”). We distinguish between three effects: Substitution, acceleration and uniqueness and find that while uniqueness emerged as having a stronger negative effect on the original’s profitability than the positive effect of acceleration, the difference between the two is relatively small. However, the negative effect of uniqueness is considerably larger than that of substitution. This is of particular interest given that industry groups have consistently concentrated on the damage caused by substitution.

Gil Appel, Barak Libai, Eitan Muller and Roni Shachar (2017), "Retention and the Monetization of Apps."
July 2017. Read paper.
Though free apps dominate mobile markets, firms struggle to monetize such products and make profits, relying on revenues from two sources: paying consumers, and paying advertisers. Accordingly, we introduce a dynamic model in which a firm offers an app in two versions: Consumers can download a free version that includes ads, or a paid version without ads. While consumers have some prior knowledge about their fit with the app, they remain uncertain about their exact match-utility unless they are using it. This match-utility drops in subsequent periods.  We show that when the drop in the match utility is low, it might be optimal to offer only the paid version. We also demonstrate that a firm can profit from offering a free version with ads even if advertisers are not paying for these ads.

Eitan Muller and Renana Peres (2017), "The Effect of Social Networks Structure on Innovation Performance:
A Review and Directions for Research."
June 2017. Read paper.
Borrowing from the field of industrial organization in economics, defined as the effect of market structure on market performance, we review the effect of social network structure on innovation performance. Specifically we discuss the effects of  (1) global characteristics of the network: average degree, degree distribution, clustering, & degree assortativity; (2) dyadic relationships: tie strength & embeddedness; (3) individual characteristics: opinion leadership & susceptibility; and (4) location in the social network: degree centrality, closeness centrality & betweenness centrality. Overall, we find that growth is particularly effective in networks that demonstrate the "3 C's": cohesive (high mutual influence among its members), connected (high number of ties), and concise (low redundancy).

back to top

working papers