1: \begin{abstract}

2: \noindent

3: This paper considers trial-offer markets where consumer preferences

4: are modeled by a multinomial logit with social influence and position

5: bias. The social signal for a product is given by its current market

6: share raised to power $r$ (\rev{or equivalently the number of purchases raised to the power of $r$}). The paper shows that, when $r$ is strictly

7: between 0 and 1, and a static position assignment (e.g., a quality

8: ranking) is used, the market converges to a unique equilibrium where

9: the market shares depend only on product quality, not their initial

10: appeals or the early dynamics. When $r$ is greater than 1, the market

11: becomes unpredictable. In many cases, the market goes to a monopoly

12: for some product: Which product becomes a monopoly depends on the

13: initial conditions of the market. These theoretical results are

14: complemented by an agent-based simulation which indicates that

15: convergence is fast when $r$ is between 0 and 1, and that the quality

16: ranking dominates the well-known popularity ranking in terms of market

17: efficiency. These results shed a new light on the role of social

18: influence which is often blamed for unpredictability, inequalities,

19: and inefficiencies in markets. In contrast, this paper shows that,

20: with a proper social signal and position assignment for the products,

21: the market becomes predictable, and inequalities and inefficiencies

22: can be controlled appropriately.

23: \end{abstract}

24: