Skip to main content
SpringerLink
Log in

Are consumers averse to sponsored messages? The role of search advertising in information discovery

  • Original Research
  • Published:
Quantitative Marketing and Economics Aims and scope Submit manuscript

Abstract

We analyze a large-scale randomized field experiment in which a search engine varied the prominence of search ads for 3.3 million US users: one group of users saw the status quo, while the other saw a lower level of advertising (with search ads sidelined). While lowering advertising significantly decreases the search engine’s revenue as expected, users exposed to the decreased level of advertising also decrease their overall search engine usage. This reduction is more significant among multi-homing users. On the supply side, going from the status quo to lower level of advertising decreases traffic to newer websites, with the newest decile losing traffic by 10%. Overall, our data suggest that viewing search ads makes consumers better off at the margin we study. We illustrate a constructive role of search advertising where advertising fills significant information gaps by conveying new information that is difficult for the search engines to gather and therefore missed by their organic algorithms.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4

Similar content being viewed by others

Data Availibility Statement

The data used in this paper’s analysis is proprietary and cannot be made available. This data belongs to an anonymous search engine and was provided to us for research purpose.

Notes

  1. Consumers conducted an estimated 20 billion search queries from US desktop computers in January 2022, with almost two-thirds of these search queries being handled by Google and most of the remaining queries being handled by Microsoft’s Bing and Oath’s Yahoo (Comscore, 2022). Among other things, consumers often search for information about current events (e.g., news, sports, celebrities), professional advice (e.g., medical, financial, legal), and products (reviews, price comparisons, new products). Examples of the variety of information sought by consumers can be seen at Google Trends. Another common type of information often sought by consumers is the location (URL) of a specific website or online service.

  2. One approach is to analyze the content of ads and document the information in the ad, such as price and product attributes (e.g., Resnik and Stern  1977; Anderson et al.  2013). Since this approach does not observe consumers’ assessments of the ads, employing it to estimate the extent of the usefulness of ads to consumers who actually see them is infeasible. Furthermore, this approach does not assess the effect of advertising on consumer information because the information consumers are exposed to in the presence and absence of ads is unobserved. Another alternative is to assess the value of viewing ads using stated preference data (e.g., Finn  1988), but this approach is limited by the low reliability of stated preferences (Aribarg et al., 2010).

  3. Our dataset includes all of the top search advertisers according to Kantar. More details on our sample and its comparison with benchmarks are presented in Section 5.

  4. Since our experiment affected a small proportion of the search engine’s user base, we assume aspects such as advertiser behavior were not affected by our experiment. Hence, our design enables the assessment of ads in the context in which they naturally appear.

  5. Buscher et al. (2010) documented that users spend ten times more attention on mainline ads versus right hand side ads. Eye-tracking studies conducted by practitioners found similar results (see Enquiro, Forbes).

  6. In this paper “query” refers to an individual search conducted by a user, characterized by the submission of specific query terms, or search phrase, to the search engine. To clarify, if a user searches for the same terms twice at two different points in time, we count them as two separate queries. Thus, the “number of queries” by a user signifies the total number of searches performed by that user on the search engine.

  7. A session is a collection of a user’s queries. A new session is initiated whenever a user conducts a new query that takes place more than 30 minutes after the last query they conducted on the search engine. Kohavi et al. (2012) discuss Bing search engine’s “overall evaluation criteria” (OEC) and mention that the number of sessions per user is a key component of it.

  8. To examine this prediction, we use cross-sectional and temporal variation in the internet presence of local businesses across the 50 states in the US, using data from US Census Bureau’s 2016 Annual Survey of Entrepreneurs (ASE).

  9. For example, unlike search advertising we study, promoted listings on many ecommerce platforms do not allow advertiser’s messages, are typically intermingled with organic listings, and sometimes significantly differ in ad allocation mechanisms. A related literature has studied the tradeoffs faced by retailers that allow search ads (e.g., Sharma and Abhishek  2017; Long et al.  2018; Yang et al.  2023).

  10. Several papers in the literature have modeled consumer interaction with TV ads and its targeting implications (e.g., Wilbur  2008; Wilbur et al.  2013; Deng and Mela  2018).

  11. One could interpret the treatment in other ways, for example, that the Low Ad group finds the search engine less aesthetically pleasing, which could affect search engine usage. It is difficult to empirically rule out such alternative interpretations without data on users’ actual views and beliefs, which is generally hard to get at scale. We support our inference using data analysis through multiple perspective, for example: Section 7.5 shows that our treatment effect depends on actual informativeness of ads and not the SERP layout; Section 8.1 shows that our treatment effect’s magnitude is highest in states where there is more local information to convey through ads, while the experimental shift in the page layout is similar across states; see also footnote 33.

  12. If both advertising and organic listings provide no value, consumers would not use the search engine.

  13. Search engines are continually updating the format and features of their SERPs. A “typical” SERP in 2022, for example, may no longer have ads on the right ad side and have different ad labels. The core of a SERP — a mainline column with a mix of sponsored search ads and organic search results, which is our focus — has largely remained the same.

  14. Constructing relevant organic listings remains a challenge for even the most established search engines. In a recent annual report, for example, Google notes that one of the major risks to their business still comes from potential deterioration of the quality of their search engine results: “...we expect web spammers will continue to seek ways to improve their rankings inappropriately. We continuously combat web spam in our search results, including through indexing technology that makes it harder for spam-like, less useful web content to rank highly...If we are subject to an increasing number of web spam, including content farms or other violations of our guidelines, this could hurt our reputation for delivering relevant information or reduce user traffic to our websites or their use of our platforms, which may adversely affect our financial condition or results.” (Alphabet, 2019)

  15. Search engines also generally offer a broad match option to advertisers that will try to find keyword additional search terms that are similar to those specified by the advertiser (a recent example of work on this topic is Grbovic et al.  2016).

  16. Similar to many web experiments, our user-level identifier is based on cookies which may not be unique to an individual. Due to this fact, we expect an attenuation bias in our estimates because all cookies in our setting are equally likely to be exposed to the treatment (Lin & Misra, 2020). Going forward in this paper we discuss cookies as users although we acknowledge that actual treatment effects may be larger.

  17. This “ad quality score” is calculated for each ad, and is distinct from the evaluations used to determine organic placement.

  18. An example of an allocation of ads can be seen in Fig. 1b, which has an allocation process similar to most other search engines. In the example, the ads with 3 highest ad scores (ads 1-3) are placed in the mainline section. This means that their ad scores are above the mainline threshold value. The ads with the next 5 highest ad scores (ads 4-8) are shown in the RHC area. The bottom ads are omitted in this example for simplicity.

  19. This sampling is a consequence of the implementation of this experiment, which achieved randomize across the experimental groups for this user segment. Due to an error in the (bucketizing) code that assigns users to experimental groups, frequent users were not correctly randomized and hence excluded from the data.

  20. Businesses are eligible for this survey if they reported more than $1,000 dollars in annual revenue. Businesses that are selected for this survey are legally required to answer the survey.

  21. The variation in the number of organic listings is caused by user settings, whether or not the search engine includes ’blended search’ style listings (e.g., maps, image results, local business listings, etc.) and a small fraction (2%) of queries that returned no search results. We count one blended search result as one organic listing.

  22. For comparison, results from Simonov and Hill (2021) imply ad CTRs between .33 to .61 for branded keywords when at least one ad is shown (depending on whether 1-4 ads are shown). In their context, this implies ad clicks account for approximately 40-70% of all clicks (i.e., at a minimum, one ad click in every three clicks).

  23. Search engines routinely pre-process user-entered query terms before performing information retrieval tasks (e.g., correcting common spelling mistakes, removing stop words, stemming words, etc.). Here we describe query terms after they have passed through this process.

  24. Results on consumer behavior in Section 7 are not limited to a single search.

  25. Our measure is based on the recency of the first visit of a website’s visitors. This is likely to be noisy for infrequently visited websites. A direct and scalable measure of a website’s “age” is not available in any dataset we know of.

  26. Dropping these missing observations is inconsequential for our takeaway from Fig. 2(c).

  27. Some examples for our data: “Online Faxing $2.99/Mo - Email, Phone & Computer Faxing”, “10 Car Insurance Quotes Online - Rates from $19 | quote .com”, “$49 Online Incorporation - Incorporate Online - 3 Easy Steps”.

  28. The US Patent and Trademark Office provides guidelines on the use of the ’\(\circledR \)’ symbol to denote registered trademarks and penalties for trademark infringement.

  29. Google Keyword Planner is a tool that allows practitioners to estimate the cost of running a search advertising campaign on Google. For any search phrase, GKP will provide the estimated price-per-click (PPC) for ads that are shown in the mainline section on Google. The tool provides a range of prices and are based on the realized prices for mainline ads over the last 12 months. Additional details are provided in Appendix H.

  30. Missing data indicates that the query terms q were either not observed frequently enough or did not have sufficient advertising over the past 12 months for Google to form an estimate. For our estimate, we assume the revenue from these ad clicks are 0.

  31. Notably, the coefficient of variation (CV) in our measurement is 35.7, representing more than a threefold increase over the standard estimate in sales contexts, where the documented CV is typically around 10 (e.g., Lewis and Rao  2015). This illustrates the inherent challenges associated with quantifying the effects of advertising on search engine usage.

  32. To make sure our results are not driven by outliers we replicate our main results after winsorizing and trimming our sample. See Appendix I.

  33. As a robustness check, we run our main analysis on the subset of users who are likely to have been exposed only to change in mainline advertising (while holding the number of RHS ads same) and find similar results, documented in Appendix E.

  34. Specifically, we identify query terms at which the number of ads change significantly going from Status Quo to Low Ad groups as ’manipulated queries’. We then examine the subset of users who searched for a ’manipulated query’ in their first search during the experimental period, as we know these users would have been affected by the experimental variation at least once. Details of this approach can be found in Appendix D.

  35. If a user-week observation is unobserved, we count it as an observation with zero queries. For instance, if a user enters the experiment on the last day of the experimental period, her usage is observed for about 4 weeks after entering the experiment (limited by our data observation window). So when we take an average, e.g., for week 12 in Fig. 3a, we count this user as making zero queries. Also note that in this figure different cohorts of users were eligible for experimental treatment for different amounts of time – those entering the experiment early got treated for more days. The analysis in this section measures effects by pooling all queries after the experiment begins.

  36. This analysis is consistent with the literature on advertising field experiments (e.g., Johnson et al.  2016) which shows that removing “unaffected” individuals from the analysis increases statistical power for estimating average effects.

  37. Note that our objective is to examine heterogeneity across individuals, and not estimate the causal effect of lowering switching costs. Further, our approach does not identify every marginal individual because individuals may reach another search engine through different means. Therefore, our test compares marginal individuals with those not necessarily on the margin.

  38. For each query term, this is the sum of squares of click-share of top three websites across all searches prior to the experiment.

  39. This measure captures the relevance-weighted variety of the organic listings. When there are a large number of listed websites, but only one is of high relevance (and is clicked), this measure would report high concentration whereas a measure based on just the number of listed websites would report low concentration.

  40. Consider a scenario where a user views three top listings on the SERP; one ad at the top followed by two organic listings in Status Quo; three organic listings in Low Ad. If organic listings all lead to the same website, removing a unique ad listing from the top will lead to the user seeing links to just one website. On the other hand, if the organic listings are diverse, removing the unique ad will add a unique organic listing to the user’s view, leading to her seeing links to the same number of unique websites.

  41. An alternative approach is to calculate OCR as the rate of occurrence of any organic click per search. Using this alternative metric gives similar results.

  42. We are assuming that the informativeness of the first SERP a consumer sees affects his belief about the search engine’s satisfying his future search needs, which affects his total future search engine usage.

  43. We do not say that conveying private information is the only motivation for advertising. Rather, we say that this mechanism drives our finding of consumers preferring advertising.

  44. In situations where the firm’s private information is verifiable by the consumer before purchase, the ad simply conveys the verifiable information. For example, in a situation where the consumer and the search engine are unaware of the firm, the ad makes the consumer aware of it. In situations where the firm’s private information cannot be verified by the consumer before purchase, for example, when the firm sells a high quality experience good, advertising serves as a costly signal (as in Sahni and Nair  2020).

  45. One could conceptualize a different model, in which there is no unknown firm, or unknown firms are inferior in the sense that they would not satisfy the consumer need. In such a model, the consumer would be weakly worse off in the presence of advertising, which is contrary to our finding. We discuss this model further in Section B.6.

References

  • - (2022). Number of explicit core search queries powered by search engines in the United States as of January 2022.https://www.statista.com/statistics/265796/us-search-engines-ranked-by-number-of-core-searches/ (visited on 03/19/2022).

  • Ackerberg, D. A. (2001). Empirically distinguishing informative and prestige effects of advertising. In: RAND Journal of economics, pp. 316–333

  • Aiello, L., et al. (2016). The role of relevance in sponsored search. In: Proceedings of the 25th ACM International on conference on information and knowledge management. ACM, pp. 185–194

  • Alphabet, Inc. (2019). 2018 Form 10-K annual report.https://abc.xyz/investor/static/pdf/20180204_alphabet_10K.pdf

  • Anderson, S. P., & Coate, S. (2005). Market provision of broadcasting: A welfare analysis. In: Review of economic studies 72.4, pp. 947–972. issn: 00346527. https://doi.org/10.1111/0034-6527.00357

  • Anderson, S. P., Ciliberto, F., & Liaukonyte, J. (2013). Information content of advertising: Empirical evidence from the OTC analgesic industry. In: International journal of industrial organization 31.5, pp. 355–367

  • Aribarg, A., Pieters R., & Wedel, M. (2010). Raising the BAR: Bias adjustment of recognition tests in advertising. In: Journal of marketing research 47.3, pp. 387–400.

  • Athey, S., & Imbens, G. W. (2017). The econometrics of randomized experiments. In: Handbook of economic field experiments. Vol. 1. Elsevier, pp. 73–140

  • Athey, S., & Nekipelov, D. (2012). A structural model of sponsored search advertising auctions. In: Working paper

  • Bisson, M. (2013). Bing ads auction explained: How bid, cost-per-click and quality score work together.https://advertise.bingads.microsoft.com/en-us/blog/post/september-2013/bing-ads-auction-explained-how-bid,-cost-perclick-and-quality-score-work-together (visited on 02/05/2019)

  • Blake, T., Nosko, C., & Tadelis, S. (2015). Consumer heterogeneity and paid search effectiveness: A Large-Scale Field Experiment. In: Econometrica 83.1, pp. 155–174. issn: 00129682. https://doi.org/10.3982/ECTA12423

  • Brin, S., & Page, L. (1998). The anatomy of a large-scale hypertextual web search engine. In: Computer networks and ISDN systems 30.1–7, pp. 107–117

  • Bronnenberg, B. J., Dubé, J.-P., & Mela, C. F. (2010). Do digital video recorders influence sales? In: Journal of marketing research 47.6, pp. 998–1010

  • Buscher, G., Dumais, S. T., & Cutrell, E. (2010). The good, the bad, and the random. In: Proceeding of the 33rd international ACM SIGIR conference on Research and development in information retrieval - SIGIR ’10. New York, New York, USA: ACM Press, p. 42. isbn: 9781450301534. https://doi.org/10.1145/1835449.1835459

  • Butters, G. R. (1977). Equilibrium distributions of sales and advertising prices. In: The Review of economic studies 44.3, pp. 465–491

  • Canel, F. (2018). bingbot Series: Maximizing crawl efficiency.https://blogs.bing.com/webmaster/october- 2018/bingbot-Series-Maximizing-Crawl-Efficiency (visited on 02/05/2019)

  • Churick, N. (2018). Google sandbox: Does google really hate new websites?https://ahrefs.com/blog/google-sandbox/ (visited on 02/05/2019)

  • Clark, C., Doraszelski, U., Draganska, M., et al. (2009). The effect of advertising on brand awareness and perceived quality: An empirical investigation using panel data. In: Quantitative marketing and economics (QME) 7.2, pp. 207–236

  • Comscore (2013). Comscore media metrix description of methodology.http://assets.pewresearch.org/wp-content/uploads/sites/13/2014/03/comScore-Media-Metrix-Description-of-Methodology.pdf (visited on 03/10/2019)

  • Dai, W., Kim, H., & Luca, M. (2023). Frontiers: Which firms gain from digital advertising? Evidence from a field experiment. In: Marketing science

  • DellaVigna, S., & Gentzkow, M. (2010). Persuasion: empirical evidence. In: Annual review of economic 2.1, pp. 643–669

  • Deng, Y., & Mela, C. F. (2018). TV viewing and advertising targeting. In: Journal of marketing research 55.1, pp. 99–118

  • DOJ (2020). Justice department sues monopolist google for violating antitrust laws. In: Justice.gov

  • Edelman, B., Ostrovsky, M., & Schwarz, M. (2007). Internet advertising and the generalized second-price auction: Selling billions of dollars worth of keywords. In: American economic review 97.1, pp. 242–259

  • eMarketer (2019). Search ad spending, US, 2019–2023.https://forecasts-na1.emarketer.com/584b26021403070290f93a59/5851918b0626310a2c186ac8 (visited on 04/08/2019)

  • Epstein, R., & Robertson, R. E. (2015). The search engine manipulation effect (SEME) and its possible impact on the outcomes of elections. In: Proceedings of the national academy of sciences 112.33, E4512–E4521

  • Erdem, T., Keane, M. P., & Sun, B. (2008). The impact of advertising on consumer price sensitivity in experience goods markets. In: Quantitative marketing and economics 6.2, pp. 139–176

  • Finn, A. (1988). Print ad recognition readership scores: An information processing perspective. In: Journal of marketing research 25.2, pp. 168–177

  • Galbraith, J. K. (1967). The new industrial state. Boston, MA: Houghton Mifflin.

    Google Scholar 

  • Goldfarb, A., & Tucker, C. (2011). Online display advertising: Targeting and obtrusiveness. In: Marketing science 30.3, pp. 389–404

  • Goldstein, D. G., et al. (2014). The economic and cognitive costs of annoying display advertisements. In: Journal of marketing research 51.6, pp. 742–752

  • Granka, L., Feusner, M., & Lorigo, L. (2008). Eye monitoring in online search. In: Passive eye monitoring. Springer, pp. 347–372

  • Grbovic, M., et al. (2016). Scalable semantic matching of queries to ads in sponsored search advertising. In: Proceedings of the 39th International ACM SIGIR conference on research and development in information retrieval. ACM, pp. 375–384

  • Grossman, G. M., & Shapiro, C. (1984). Informative advertising with differentiated products. In: The Review of economic studies 51.1, pp. 63–81

  • Haahr, P. (2016). SMX West 2016 - how google works: A google ranking engineer’s story. https://www.youtube.com/watch?v=iJPu4vHETXw (visited on 02/05/2019).

  • Huang, J., Reiley, D., & Riabov, N. (2018). Measuring consumer sensitivity to audio advertising: A field experiment on pandora internet radio. In.

  • Huang, J., White, R. W., & Dumais, S. (2011). No clicks, no problem: using cursor movements to understand and improve search. In: Proceedings of the SIGCHI conference on human factors in computing systems, pp. 1225–1234

  • Jeziorski, P., & Segal, I. (2015). What makes them click: Empirical analysis of consumer demand for search advertising. In: American economic journal: Microeconomics 7.3, pp. 24–53

  • Johnson, J. P. (2013). Targeted advertising and advertising avoidance. In: The RAND journal of economics 44.1, pp. 128–144

  • Johnson, G. A. (2022). Inferno: A guide to field experiments in online display advertising. In: Journal of economics & management strategy

  • Johnson, G. A., Lewis, R. A., & Reiley, D. H. (2016). When less is more: Data and power in advertising experiments. In: Marketing science 36.1, pp. 43–53

  • Kaiser, U., & Song, M. (2009). Do media consumers really dislike advertising? An empirical assessment of the role of advertising in print media markets. In: International journal of industrial organization 27.2, pp. 292–301

  • Kaldor, N. (1950). The economic aspects of advertising. In: The Review of economic studies 18.1, pp. 1–27. issn: 00346527, 1467937X. http://www.jstor.org/stable/2296103

  • Kenshoo (2020). Search advertising clickthrough rate (CTR) worldwide from 2nd quarter 2018 to 2nd quarter of 2020.https://www.statista.com/statistics/873637/search-advertising-ctr/ (visited on 03/07/2022)

  • Kihlstrom, R. E., & Riordan, M. H. (1984). Advertising as a signal. In: Journal of political economy 92.3, pp. 427–450

  • Kohavi, R., et al. (2012). Trustworthy online controlled experiments: Five puzzling outcomes explained. In: Proceedings of the 18th ACM SIGKDD international conference on knowledge discovery and data mining. ACM, pp. 786–794

  • Lewis, R. A., & Rao, J. M. (2015). The unfavorable economics of measuring the returns to advertising. In: The quarterly journal of economics 130.4, pp. 1941–1973

  • Lin, T., & Misra, S. (2020). The identity fragmentation bias. In: arXiv preprintarXiv:2008.12849

  • Long, F., Jerath, K., Sarvary, M. (2018). The informational role of sponsored advertising on online retail marketplaces. Technical report

  • McDevitt, R. C. (2014). ”A” business by any other name: Firm name choice as a signal of firm quality. In: Journal of political economy 122.4, pp. 909–944

  • Mehta, N., Chen, X., & Narasimhan, O. (2008). Informing, transforming, and persuading: Disentangling the multiple effects of advertising on brand choice decisions. In: Marketing science 27.3, pp. 334–355

  • Milgrom, P., & Roberts, J. (1986). Price and advertising signals of product quality. In: Journal of political economy 94.4, pp. 796–821

  • Moshary, S. (2021). Sponsored search in equilibrium: Evidence from two experiments. In: Working paper

  • Mueller, J. (2018). How often does google re-index websites?https://www.youtube.com/watch?v=5pH2HpoNk7k (visited on 02/05/2019)

  • Narayanan, S., & Kalyanam, K. (2015). Position effects in search advertising and their moderators: A regression discontinuity approach. In: Marketing science 34.3, pp. 388–407

  • Nelson, P. (1974). Advertising as information. In: Journal of political economy 82.4, pp. 729–754

  • Rayo, L., & Segal, I. (2010). Optimal information disclosure. In: Journal of political economy 118.5, pp. 949–987

  • Reiley, D. H., Li, S.-M., & Lewis, R. A. (2010). Northern exposure: A field experiment measuring externalities between search advertisements. In: Proceedings of the 11th ACM conference on electronic commerce. ACM, pp. 297–304

  • Resnik, A., & Stern, B. L. (1977). An analysis of information content in television advertising. In: Journal of marketing 41.1, pp. 50–53

  • Robinson, J. (1933). Economics of imperfect competition

  • Rysman, M. (2004). Competition between networks: A study of the market for yellow pages. In: The Review of economic studies 71.2, pp. 483–512

  • Sahni, N. S., & Nair, H. S. (2020). Does advertising serve as a signal? Evidence from a field experiment in mobile search. In: The review of economic studies 87.3, pp. 1529–1564

  • Schwartz, B. (2016). Google’s search knows about over 130 trillion pages.https://searchengineland.com/googlessearch-indexes-hits-130-trillion-pages-documents-263378 (visited on 02/15/2019)

  • Sharma, S., & Abhishek, V. (2017). Effect of sponsored listings on online marketplaces: The role of information asymmetry. In: Available at SSRN 3013468

  • Siddarth, S., & Chattopadhyay, A. (1998). To zap or not to zap: A study of the determinants of channel switching during commercials. In: Marketing science 17.2, pp. 124–138

  • Simonov, A., & Hill, S. (2021). Competitive advertising on brand search: Traffic stealing and click quality. In: Marketing science 40.5, pp. 923–945

  • Simonov, A., Nosko, C., & Rao, J. M. (2018). Competition and crowd-out for brand keywords in sponsored search. In: Marketing science 37.2, pp. 200–215. issn: 0732-2399. https://doi.org/10.1287/mksc.2017.1065

  • Stigler, G. J. (1961). The economics of information. In: Journal of political economy 69.3, pp. 213–225

  • Stigler, G. J., & Becker, G. S. (1977). De gustibus non est disputandum. In: American economic review 67.2, pp. 76–90

  • Tellis, G. J., Chandy, R. K., & Thaivanich, P. (2000). Which ad works, when, where, and how often? Modeling the effects of direct television advertising. In: Journal of marketing research 37.1, pp. 32–46

  • Telser, L. G. (1964). Advertising and competition. In: Journal of political economy 72.6, pp. 537–562

  • Tuchman, A. E., Nair, H., & Gardete, P. M. (2018). Television ad-skipping, consumption complementarities and the consumer demand for advertising. In: Quantitative marketing and economics, pp. 1–64

  • Tucker, C. E. (2012). The economics of advertising and privacy. In: International journal of industrial organization 30.3, pp. 326–329

  • Wilbur, K. C. (2008). A two-sided, empirical model of television advertising and viewing markets. In: Marketing science 27.3, pp. 356–378

  • Wilbur, K. C., Xu, L., & Kempe, D. (2013). Correcting audience externalities in television advertising. In: Marketing science 32.6, pp. 892–912

  • Yang, J., et al. (2023). Advertising as information for ranking e-commerce search listings. In: Marketing science

  • Yao, S., & Mela, C. F. (2011). A dynamic model of sponsored search advertising. In: Marketing science 30.3, pp. 447–468

  • Yin, D., et al. (2016). Ranking relevance in yahoo search. In: Proceedings of the 22nd ACM SIGKDD International conference on knowledge discovery and data mining. ACM, pp. 323–332

Download references

Author information

Authors and Affiliations

  1. Stanford GSB, Stanford, CA, USA

    Navdeep S. Sahni & Charles Zhang

Authors
  1. Navdeep S. Sahni
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Charles Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Navdeep S. Sahni.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Previous drafts: Aug 23 2019; October 2019; June 2020; October 2020; Feb 2021; July 2021; Nov 2021; July 2022. The views discussed here represent that of the authors and not of Stanford University. We are thankful to Thomas Otter and anonymous reviewers at QME for their feedback. We are also thankful for comments from Pedro Gardete, Gagan Goel, Sachin Gupta, Wes Hartmann, Carl Mela, Sridhar Moorthy, Harikesh Nair, Sridhar Narayanan, Devesh Raval, Stephan Seiler, Andrey Simonov, Raluca Ursu, Caio Waisman, Ali Yurukoglu, Marcos Salgado and from participants at the 2021 NBER IO Winter Meetings, AFE Conference University of Chicago, MIT Conference on Digital Experimentation, CMU Digital Marketing and ML conference, 2019 Choice Symposium, 2019 & 2020 Marketing Science conference, Marketplaces and Algorithms Design seminar 2020, 2020 INFORMS Conference; seminars at Wharton, Yale SOM, Google, Washington Univ, UC Riverside, Harvard Business School, UIUC, SMU, Univ of Florida, Santa Clara Univ. The usual disclaimer applies. Please contact Sahni (navdeep.sahni@stanford.edu) or Zhang (cyzhang@stanford.edu) for correspondence.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary file 1 (pdf 1582 KB)

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Sahni, N.S., Zhang, C. Are consumers averse to sponsored messages? The role of search advertising in information discovery. Quant Mark Econ 22, 63–114 (2024). https://doi.org/10.1007/s11129-023-09270-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11129-023-09270-z

Keywords

JEL Classification

Search

Navigation