Results 1 
6 of
6
Revenue Optimization in the Generalized SecondPrice Auction
"... We consider the optimization of revenue in advertising auctions based on the generalized secondprice (GSP) paradigm, which has become a de facto standard. We examine several different GSP variants (including squashing and different types of reserve prices), and consider how to set their parameters ..."
Abstract

Cited by 8 (0 self)
 Add to MetaCart
We consider the optimization of revenue in advertising auctions based on the generalized secondprice (GSP) paradigm, which has become a de facto standard. We examine several different GSP variants (including squashing and different types of reserve prices), and consider how to set their parameters optimally. One intriguing finding is that charging each advertiser the same perclick reserve price (“unweighted reserve prices”) yields dramatically more revenue than the qualityweighted reserve prices that have become common practice. This result is robust, arising both from theoretical analysis and from two different kinds of computational experiments. We also identify a new GSP variant that is revenue optimal in restricted settings. Finally, we study how squashing and reserve prices interact, and how equilibrium selection affects the revenue of GSP when features such as reserves or squashing are applied.
NearOptimal MultiUnit Auctions with Ordered Bidders
, 2013
"... We construct priorfree auctions with constantfactor approximation guarantees with ordered bidders, in both unlimited and limited supply settings. We compare the expected revenue of our auctions on a bid vector to the monotone price benchmark, the maximum revenue that can be obtained from a bid vec ..."
Abstract

Cited by 2 (1 self)
 Add to MetaCart
(Show Context)
We construct priorfree auctions with constantfactor approximation guarantees with ordered bidders, in both unlimited and limited supply settings. We compare the expected revenue of our auctions on a bid vector to the monotone price benchmark, the maximum revenue that can be obtained from a bid vector using supplyrespecting prices that are nonincreasing in the bidder ordering and bounded above by the secondhighest bid. As a consequence, our auctions are simultaneously nearoptimal in a wide range of Bayesian multiunit environments.
Selling in Exclusive Markets: Some Observations on Priorfree Mechanism Design
, 2012
"... We consider priorfree benchmarks in nonmatroid settings. In particular, we show that a very desirable benchmark proposed by Hartline and Roughgarden [22] is too strong, in the sense that no truthful mechanism can compete with it even in a very simple nonmatroid setting where there are two exclusi ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
We consider priorfree benchmarks in nonmatroid settings. In particular, we show that a very desirable benchmark proposed by Hartline and Roughgarden [22] is too strong, in the sense that no truthful mechanism can compete with it even in a very simple nonmatroid setting where there are two exclusive markets and the seller can only sell to agents in one of them. On the other hand, we show that there is a mechanism that competes with a symmetrized version of this benchmark. We further investigate the more traditional best fixed price profit benchmark and show that there are mechanisms that compete with it in any downwardclosed settings.
DESIGN AND ANALYSIS OF SPONSORED SEARCH MECHANISMS
, 2012
"... Auctions have become the standard way of allocating resources in electronic markets. Two main reasons why designing auctions is hard are the need to cope with strategic behavior of the agents, who will constantly adjust their bids seeking more items at lower prices, and the fact that the environment ..."
Abstract
 Add to MetaCart
Auctions have become the standard way of allocating resources in electronic markets. Two main reasons why designing auctions is hard are the need to cope with strategic behavior of the agents, who will constantly adjust their bids seeking more items at lower prices, and the fact that the environment is highly dynamic and uncertain. Many market designs which became defacto industrial standards allow strategic manipulation by the agents, but nevertheless display good behavior in practice. In this thesis, we analyze why such designs turned out to be so successful despite strategic behavior and environment uncertainty. Our goal is to learn from this analysis and to use the lessons learned to design new auction mechanisms; as well as finetune the existing ones. We illustrate this research line through the analysis and design of Ad Auctions mechanisms. We do so by studying the equilibrium behavior of a game induced by Ad Auctions, and show that all equilibria have good welfare and revenue properties. Next, we present new Ad Auction designs that take into account
SOCIAL STATUS AND BADGE DESIGN
"... Many websites encourage user participation via the use of virtual rewards like badges. While badges typically have no explicit value, they act as symbols of social status within a community. In this paper, we study how to design virtual incentive mechanisms that maximize total contributions made t ..."
Abstract
 Add to MetaCart
(Show Context)
Many websites encourage user participation via the use of virtual rewards like badges. While badges typically have no explicit value, they act as symbols of social status within a community. In this paper, we study how to design virtual incentive mechanisms that maximize total contributions made to a website when badges are only valued as a symbol of social status. We consider a gametheoretic model where users exert costly effort to make contributions and, in return, are awarded with badges. The value of a badge is determined endogenously by the number of users who earn an equal or higher badge; as more users earn a particular badge, the value of that badge diminishes for all users. We show that among all possible mechanisms for assigning statusdriven rewards, the optimal mechanism is a leaderboard with a cutoff: users that contribute less than a certain threshold receive nothing while the remaining are ranked by contribution. We next study the necessary features of approximately optimal mechanisms and find that approximate optimality is influenced by the convexity of status valuations, i.e. whether being ranked above more people has an increasing or decreasing marginal effect in a user’s satisfaction. When status valuations are concave, any approximately optimal mechanism must contain a coarse status partition, i.e. a partition of users in status classes whose size will grow as the number of users grows. Conversely when status valuations are convex, we prove that fine partitioning, i.e. a partition of users in status classes whose size stays constant as the number of users grow, is necessary for approximate optimality.
PriorFree MultiUnit Auctions with Ordered Bidders
, 2013
"... Priorfree auctions are robust auctions that assume no distribution over bidders’ valuations and provide worstcase (inputbyinput) approximation guarantees. In contrast to previous work on this topic, we pursue good priorfree auctions with nonidentical bidders. Priorfree auctions can approximat ..."
Abstract
 Add to MetaCart
(Show Context)
Priorfree auctions are robust auctions that assume no distribution over bidders’ valuations and provide worstcase (inputbyinput) approximation guarantees. In contrast to previous work on this topic, we pursue good priorfree auctions with nonidentical bidders. Priorfree auctions can approximate meaningful benchmarks for nonidentical bidders only when sufficient qualitative information about the bidder asymmetry is publicly known. We consider digital goods auctions where there is a total ordering of the bidders that is known to the seller, where earlier bidders are in some sense thought to have higher valuations. We use the framework of Hartline and Roughgarden (STOC ’08) to define an appropriate revenue benchmark: the maximum revenue that can be obtained from a bid vector using prices that are nonincreasing in the bidder ordering and bounded above by the secondhighest bid. This monotoneprice benchmark is always as large as the wellknown fixedprice benchmark F (2), so designing priorfree auctions with good approximation guarantees is only harder. By design, an auction that approximates the monotoneprice benchmark satisfies a very strong guarantee: it is, in particular, simultaneously nearoptimal for essentially every Bayesian environment in