What Makes Them Click: Empirical Analysis of Consumer Demand for Search Advertising

Jeziorski, Przemysław; Segal, Ilya

doi:10.2139/ssrn.1417625

Cited by 70 publications

(52 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A myriad of researchers have turned to SSA data to uncover new insights about consumer search (Ghose and Yang, 2009;Rutz and Bucklin, 2011), choice and related purchasing behaviors (Jeziorski and Segal, 2009;Yang and Ghose, 2010;Agarwal et al, 2011) and advertiser/search engine strategies (Animesh et al, 2009;Yao and Mela, 2011;Rutz et al, 2012). Many of these papers have used random utility models to study the effect of ad position, keyword length, presence or absence of brand name, etc.…”

Section: Introductionmentioning

confidence: 99%

Aggregation Bias in Sponsored Search Data: The Curse and the Cure

2015

View full text Add to dashboard Cite

There has been significant recent interest in studying consumer behavior in sponsored search advertising (SSA). Researchers have typically used daily data from search engines containing measures such as average bid, average ad position, total impressions, clicks and cost for each keyword in the advertiser's campaign. A variety of random utility models have been estimated using such data and the results have helped researchers explore the factors that drive consumer click and conversion propensities. However, virtually every analysis of this kind has ignored the intra-day variation in ad position. We show that estimating random utility models on aggregated (daily) data without accounting for this variation will lead to systematically biased estimates -specifically, the impact of ad position on click-through rate (CTR) is attenuated and the predicted CTR is higher than the actual CTR. We demonstrate the existence of the bias analytically and show the effect of the bias on the equilibrium of the SSA auction. Using a large dataset from a major search engine, we measure the magnitude of bias and quantify the losses suffered by the search engine and an advertiser using aggregate data. The search engine revenue loss can be as high as 11% due to aggregation bias. We also present a few data summarization techniques that can be used by search engines to reduce or eliminate the bias.

show abstract

Section: Introductionmentioning

confidence: 99%

Aggregation Bias in Sponsored Search Data: The Curse and the Cure

2015

View full text Add to dashboard Cite

show abstract

“…Rohlfs [16] discusses positive externalities in the telephone industry in which a person's value for a telephone increases as more friends use it. A well-studied scenario with negative externalities is the allocation of ad slots in which, a company's valuation for being listed as one of the sponsored search results decreases if their competitor is also listed [1,4,9,10,13,14]. Finally, the valuation might have mixed externalities, as in the sale of nuclear weapons [12], in which countries prefer their allies rather than their foes to win the auction.…”

Section: Introductionmentioning

confidence: 99%

Optimal auctions with positive network externalities

Haghpanah

Immorlica

Mirrokni

et al. 2011

Proceedings of the 12th ACM Conference on Electronic Commerce

View full text Add to dashboard Cite

We consider the problem of designing auctions in social networks for goods that exhibit single-parameter submodular network externalities in which a bidder's value for an outcome is a fixed private type times a known submodular function of the allocation of his friends. Externalities pose many issues that are hard to address with traditional techniques; our work shows how to resolve these issues in a specific setting of particular interest. We operate in a Bayesian environment and so assume private values are drawn according to known distributions. We prove that the optimal auction is APX-hard. Thus we instead design auctions whose revenue approximates that of the optimal auction. Our main result considers step-function externalities in which a bidder's value for an outcome is either zero, or equal to his private type if at least one friend has the good. For these settings, we provide a e e+1 -approximation. We also give a 0.25-approximation auction for general single-parameter submodular network externalities, and discuss optimizing over a class of simple pricing strategies.

show abstract

“…(There has been significant recent progress on inferring valuation distributions from bids or other indications of demand in a variety of applications [17,18,21,26], but that work typically used bids in auctions or survey information. )…”

Section: Estimating a Rich Customer Valuation Modelmentioning

confidence: 99%

A Framework for Automated Bundling and Pricing Using Purchase Data

Benisch

Sandholm

2012

Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering

View full text Add to dashboard Cite

Abstract. We present a framework for automatically suggesting highprofit bundle discounts based on historical customer purchase data. We develop several search algorithms that identify profit-maximizing prices and bundle discounts. We introduce a richer probabilistic valuation model than prior work by capturing complementarity, substitutability, and covariance, and we provide a hybrid search technique for fitting such a model to historical shopping cart data. As new purchase data is collected, it is integrated into the valuation model, leading to an online technique that continually refines prices and bundle discounts. To our knowledge, this is the first paper to study bundle discounting using shopping cart data. We conduct computational experiments using our fitting and pricing algorithms that demonstrate several conditions under which offering discounts on bundles can benefit the seller, the buyer, and the economy as a whole. One of our main findings is that, in contrast to products typically suggested by recommender systems, the most profitable products to offer bundle discounts on appear to be those that are occasionally purchased together and often separately.

show abstract

What Makes Them Click: Empirical Analysis of Consumer Demand for Search Advertising

Cited by 70 publications

References 22 publications

Aggregation Bias in Sponsored Search Data: The Curse and the Cure

Aggregation Bias in Sponsored Search Data: The Curse and the Cure

Optimal auctions with positive network externalities

A Framework for Automated Bundling and Pricing Using Purchase Data

Contact Info

Product

Resources

About