Ads (Link Bibliography)

“Ads” links:

  1. AB-testing








  9. Milk

  10. Candy-Japan

  11. Traffic#adwords





  16. 2014-goldstein.pdf: ⁠, Daniel G. Goldstein, Siddharth Suri, R. Preston McAfee, Matthew Ekstrand-Abueg, Fernando Díaz (2014; advertising⁠, economics⁠, technology⁠, psychology):

    Some online display advertisements are annoying. Although publishers know the payment they receive to run annoying ads, little is known about the cost that such ads incur (eg., causing website abandonment). Across three empirical studies, the authors address two primary questions: (1) What is the economic cost of annoying ads to publishers? and (2) What is the cognitive impact of annoying ads to users? First, the authors conduct a preliminary study to identify sets of more and less annoying ads. Second, in a field experiment, they calculate the compensating differential, that is, the amount of money a publisher would need to pay users to generate the same number of impressions in the presence of annoying ads as it would generate in their absence. Third, the authors conduct a mouse-tracking study to investigate how annoying ads affect reading processes. They conclude that in plausible scenarios, the practice of running annoying ads can cost more money than it earns.

  17. #mccoy-et-al-2007-section



  20. ⁠, Anna Kerkhof (2019-09-04; economics⁠, advertising):

    Does advertising revenue increase or diminish content differentiation in media markets? This paper shows that an increase in the technically feasible number of ad breaks per video leads to an increase in content differentiation between several thousand YouTube channels. I exploit two institutional features of YouTube’s monetization policy to identify the causal effect of advertising on the YouTubers’ content choice. The analysis of around one million YouTube videos shows that advertising leads to a twenty percentage point reduction in the YouTubers’ probability to duplicate popular content, ie., content in high demand by the audience. I also provide evidence of the economic mechanism behind the result: popular content is covered by many competing YouTubers; hence, viewers who perceive advertising as a nuisance could easily switch to a competitor if a YouTuber increased her number of ad-breaks per video. This is less likely, however, when the YouTuber differentiates her content from her competitors.

    [Keywords: advertising, content differentiation, economics of digitization, horizontal product differentiation, long tail, media diversity, user-generated content, YouTube]

    …The analysis of around one million YouTube videos shows that an increase in the feasible number of ad breaks per video leads to a twenty percentage point reduction in the YouTubers’ probability to duplicate popular content. The is considerable: it corresponds to around 40% of a standard deviation in the dependent variable and to around 50% of its baseline value.

    The large sample size allows me to conduct several sub-group analyses to study effect heterogeneity. I find that the positive effect of advertising on content differentiation is driven by the YouTubers who have at least 1,000 subscribers, ie., the YouTubers whose additional ad revenue is likely to exceed the costs from adapt-ing their videos’ content. In addition, I find heterogeneity along video categories: some categories are more flexible in terms of their typical video duration than others, hence, exploiting the ten minutes trick is more easy (eg., a music clip is typically between three and five minutes long and cannot be easily extended). A battery of robustness checks confirms these results.

    …Moreover, I show that ad revenue does not necessarily improve the YouTubers’ video quality. Although the number of views goes up when a video has more ad breaks, the relative number of likes decreases…Table 5 shows the results. The size of the estimates for δ′′(columns 1 to 3), though at the 1%-level, is negligible: a one second increase in video duration corresponds to a 0.0001 percentage point increase in the fraction of likes. The estimates for δ′′′ in columns 4 to 6, though, are relatively large and statistically-significant at the 1%-level, too. According to these estimates, one further second in video duration leads on average to about 1.5 percent more views. These estimates may reflect the algorithmic drift discussed in Section 9.2. YouTube wants to keep its viewers as long as possible on the platform to show as many ads as possible to them. As a result, longer videos get higher rankings and are watched more often.

  21. ⁠, Seth G. Benzell, Avinash Collis (2019-10-12; economics⁠, advertising):

    Digital platforms, such as Facebook, ⁠, and AirBnB, create value by connecting users, creators, and contractors of different types. Their rapid growth, untraditional business model, and disruptive nature presents challenges for managers and asset pricers. These features also, arguably, make them natural monopolies, leading to increasing calls for special regulations and taxes. We construct and illustrate a approach for modeling digital platforms. The model allows for heterogeneity in elasticity of demand and heterogeneous network effects across different users. We parameterize our model using a survey of over 40,000 US internet users on their demand for Facebook. Facebook creates about 11.2 billion dollars in consumer surplus a month for US users age 25 or over, in line with previous estimates. We find Facebook has too low a level of advertising relative to their revenue maximizing strategy, suggesting that they also value maintaining a large user base. We simulate six proposed government policies for digital platforms, taking Facebook’s optimal response into account. Taxes only slightly change consumer surplus. Three more radical proposals, including ‘data as labor’ and nationalization, have the potential to raise consumer surplus by up to 42%. But a botched regulation that left the US with two smaller, non-competitive social media monopolies would decrease consumer surplus by 44%.

  22. 2017-sinha.pdf: ⁠, Atanu R. Sinha, Meghanath Macha, Pranav Maneriker, Sopan Khosla, Avani Samdariya, Navjot Singh (2017-08-14; advertising⁠, technology):

    The increasing use of ad blocking software poses a major threat for publishers in loss of online ad revenue, and for advertisers in the loss of audience. Major publishers have adopted various anti-ad blocking strategies such as denial of access to website content and asking users to subscribe to paid ad-free versions. However, publishers are unsure about the true impact of these strategies2, 3. We posit that the real problem lies in the measurement of effectiveness because the existing methods compare metrics after implementation of such strategies with that of metrics just before implementation, making them error prone due to sampling bias. The errors arise due to differences in group compositions across before and after periods, as well as differences in time-period selection for the before measurement. We propose a novel algorithmic method which modifies the difference-in-differences approach to address the sampling bias due to differences in time-period selection. Unlike difference-in-differences, we choose the time-period for comparison in an endogenous manner, as well as, exploit differences in ad blocking tendencies among visitors’ arriving on the publisher’s site to allow cluster specific choice of the control time-period. We evaluate the method on both synthetic data (which we make available) and proprietary real data from an online publisher and find good support.

  23. 2017-sinha-figure3-peruserantiadblockeffects.png

  24. 1991-abernethy.pdf: ⁠, Avery M. Abernethy (1991; advertising⁠, economics):

    Although it is generally accepted that television program ratings are greater than the audience’s exposure to the advertising, the key issue is the actual size of the difference. A review of advertising, marketing, communication, and sociology literature yields some indications of the degree of difference between ad and program exposure and factors in the viewing environment which could influence audience commercial avoidance.

  25. 2000-bayles-justhowblindarewetoadvertisingbannersontheweb.html: ⁠, Michelle Bayles (2000-07; advertising⁠, technology⁠, psychology):

    Moreover, Benway (1998) showed that extremely colorful and obvious banners tend to be ignored by users. When participants in this study were asked to find specific information on a web page, the information was not found if it was embedded in a banner. Benway consequently named this phenomenon “banner blindness.” Benway also found that banners located at the top of the page (away from other links), tended to be ignored more often than banners located lower down the page (closer to other important links). This finding is supported by another study which showed a 77% increased click-through rate for advertisements placed 1⁄3 of the way down the page (Athenia Assoc., 1997).

    …In our study, we were curious to simply explore how much users remember about a web page after viewing it—in particular, we were interested in investigating user memory of banner advertisements:

    1. How well can users’ recall a banner advertisement on a web page?
    2. How well can users’ recognize a banner advertisement on a web page?
    3. Does animation affect user recall or recognition of an advertising banner?

    …Very few participants were able to complete both the recognition and recall tasks correctly. Only 3 (9%) of participants were able to correctly recall both advertisements, recognize both companies, and correctly recall and recognize the state in which they were presented. On the other hand, participants who were unable to recall anything for either company banner or correctly indicate the animation state of the banner (40%) had a surprisingly high recognition rate of 79% for two correctly recognized ads. Results also show that of the 26% who recognized only one ad, the banner recognized was typically presented in the animated state. In other words, 7 out of 9 times the single banner correctly recognized was in the animated state. This indicates that animation may have some effect on recognition.

    Results from this study indicate that recognition of the banner advertisements were fairly high (74% for both banners). In addition, about half of the participants were able to recall at least seeing an advertisement on the page—and many of these actually recalled the name of the company. These results show that most users did notice and remember the banners even though they were not part of the search tasks they were performing.

  26. 2002-edwards.pdf: ⁠, Steven M. Edwards, Hairong Li, Joo-Hyun Lee (2002; advertising):

    This paper explores forced viewing of “pop-up ads” on the Internet to understand better how viewers come to define ads as irritating and decide to avoid them. Perceived intrusiveness was suggested as the underlying mechanism by which the process occurs. Antecedents of intrusiveness were identified that affect perceptions of ads as interruptions, including congruence of the advertisement content with the current task and intensity of cognition at the moment the ad pops up. The consequences of intrusiveness were shown to be caused by feelings of irritation and ad avoidance. The results provide an understanding of how consumers experience forced exposure situations in interactive environments and highlight implications for advertisers seeking to increase the effectiveness of on-line advertising.



  29. 2020-michelon.pdf: ⁠, Aaron Michelon, Steven Bellman, Margaret Faulkner, Justin Cohen, Johan Bruwer (2020-03-23; advertising⁠, economics⁠, technology):

    Radio remains popular, delivering an audience reach of over 90 percent, but radio ratings may overestimate real advertising exposure. Little is known about audience and media factors affecting radio-advertising avoidance. Many advertisers have believed as much as one-third of the audience switch stations during radio-advertising breaks. In the current study, the authors combined Canadian portable people-meter data ratings to measure loss of audience during advertising. They discovered a new benchmark of 3% (across conditions) for mechanical (or actual physical) avoidance of radio advertising, such as switching stations or turning off the radio. This rate is about one-tenth of current estimates, but was higher for music versus talk stations, out-of-home versus in-home listening, and early versus late dayparts.

  30. #google

  31. DNB-FAQ



  34. 2018-berman.pdf: “p-Hacking and False Discovery in A  /​ ​​ ​B Testing”⁠, Ron Berman, Leonid Pekelis, Aisling Scott, Christophe Van den Bulte


  36. #pandora

  37. #mozilla

  38. #suarez-garcia-marinoso-2021

  39. #linkedin

  40. ⁠, Ron Kohavi, Alex Deng, Brian Frasca, Roger Longbotham, Toby Walker, Ya Xu (2012-08-12):

    Online controlled experiments are often utilized to make data-driven decisions at Amazon, Microsoft, eBay, Facebook, Google, Yahoo, Zynga, and at many other companies. While the theory of a is simple, and dates back to Sir Ronald A. Fisher’s experiments at the Rothamsted Agricultural Experimental Station in England in the 1920s, the deployment and mining of online controlled experiments at scale—thousands of experiments now—has taught us many lessons. These exemplify the proverb that the difference between theory and practice is greater in practice than in theory. We present our learnings as they happened: puzzling outcomes of controlled experiments that we analyzed deeply to understand and explain. Each of these took multiple-person weeks to months to properly analyze and get to the often surprising root cause. The root causes behind these puzzling results are not isolated incidents; these issues generalized to multiple experiments. The heightened awareness should help readers increase the trustworthiness of the results coming out of controlled experiments. At Microsoft’s Bing, it is not uncommon to see experiments that impact annual revenue by millions of dollars, thus getting trustworthy results is critical and investing in understanding anomalies has tremendous payoff: reversing a single incorrect decision based on the results of an experiment can fund a whole team of analysts. The topics we cover include: the OEC (Overall Evaluation Criterion), click tracking, effect trends, experiment length and power, and carryover effects.

  41. #pagefair

  42. #yan-et-al-2020

  43. #aral-dhillon-2020

  44. ⁠, Jason Huang, David H. Reiley, Nickolai M. Riabov (2018-04-21; advertising):

    A randomized experiment with almost 35 million Pandora listeners enables us to measure the sensitivity of consumers to advertising, an important topic of study in the era of ad-supported digital content provision. The experiment randomized listeners into 9 treatment groups, each of which received a different level of audio advertising interrupting their music listening, with the highest treatment group receiving more than twice as many ads as the lowest treatment group. By keeping consistent treatment assignment for 21 months, we are able to measure long-run demand effects, with three times as much ad-load sensitivity as we would have obtained if we had run a month-long experiment. We estimate a demand curve that is strikingly linear, with the number of hours listened decreasing linearly in the number of ads per hour (also known as the price of ad-supported listening). We also show the negative impact on the number of days listened and on the probability of listening at all in the final month. Using an experimental design that separately varies the number of commercial interruptions per hour and the number of ads per commercial interruption, we find that neither makes much difference to listeners beyond their impact on the total number of ads per hour. Lastly, we find that increased ad load causes a substantial increase in the number of paid ad-free subscriptions to Pandora, particularly among older listeners.


  46. ⁠, Ben Miroglio, David Zeber, Jofish Kaye, Rebecca Weiss (2018-04-23):

    Web users are increasingly turning to ad blockers to avoid ads, which are often perceived as annoying or an invasion of privacy. While there has been substantial research into the factors driving ad blocker adoption and the detrimental effect to ad publishers on the Web, the resulting effects of ad blocker usage on Web users’’ browsing experience is not well understood. To approach this problem, we conduct a retrospective natural field experiment using Firefox browser usage data, with the goal of estimating the effect of adblocking on user engagement with the Web. We focus on new users who installed an ad blocker after a baseline observation period, to avoid comparing different populations. Their subsequent browser activity is compared against that of a control group, whose members do not use ad blockers, over a corresponding observation period, controlling for baseline usage. In order to estimate causal effects, we employ propensity score matching on a number of other features recorded during the baseline period. In the group that installed an ad blocker, we find substantial increases in both active time spent in the browser (+28% over control) and the number of pages viewed (+15% over control), while seeing no change in the number of searches. Additionally, by reapplying the same methodology to other popular Firefox browser extensions, we show that these effects are specific to ad blockers. We conclude that ad blocking has a positive impact on user engagement with the Web, suggesting that any costs of using ad blockers to users’’ browsing experience are largely drowned out by the utility that they offer.

  47. Everything

  48. #gordon-et-al-2019-2

  49. ⁠, Jinyun Yan, Birjodh Tiwana, Souvik Ghosh, Haishan Liu, Shaunak Chatterjee (2019-01-29):

    Organic updates (from a member’s network) and sponsored updates (or ads, from advertisers) together form the newsfeed on LinkedIn. The newsfeed, the default homepage for members, attracts them to engage, brings them value and helps LinkedIn grow. Engagement and Revenue on feed are two critical, yet often conflicting objectives. Hence, it is important to design a good Revenue-Engagement Tradeoff (RENT) mechanism to blend ads in the feed. In this paper, we design experiments to understand how members’ behavior evolve over time given different ads experiences. These experiences vary on ads density, while the quality of ads (ensured by relevance models) is held constant. Our experiments have been conducted on randomized member buckets and we use two experimental designs to measure the short term and long term effects of the various treatments. Based on the first three months’ data, we observe that the long term impact is at a much smaller scale than the short term impact in our application. Furthermore, we observe different member cohorts (based on user activity level) adapt and react differently over time.

  50. 2019-yan-linkedin-figure3-userharm.png

  51. 2007-mccoy.pdf: ⁠, Scott McCoy, Andrea Everard, Peter Polak, Dennis F. Galletta (2007; advertising⁠, technology⁠, psychology):

    We conducted an experiment with different forms and types of ads. An artificial Web site was created for the experiment that contained images, prices, and descriptions of familiar products and product categories. The products were those that would be carried by a general store and included food, health care, and household products. Nine search tasks were assigned to participants that would force them to traverse a variety of portions of the site…The experimental websites were accessed over the Internet in a controlled laboratory setting by 536 undergraduate students.

    …This study provides clear support for an assertion that users will adopt more negative intentions when a site displays advertisements than when the site does not. It is also clear that advertisements interfere with retention of site content and that features of advertisements also have important effects on retaining both site and ad content. Inline ads permit both site and ad content to be remembered more clearly than popups and popunders, a finding that is most interesting because it suggests the action of closing the advertisement window distracts users from the site, and further, it is visible for a shorter time. When ads are markedly different from the content of the site, they theoretically stimulate more effort as users work toward an important goal, and users remember more about both the Web site and the advertisement. It is interesting to note that while these effects might on the surface appear small, they are quite consistent and highly statistically-significant. Extrapolating to millions of site visitors, even small differences can amount to an urgent problem for management. Finally, it is also clear that popups and popunders are considered to be more intrusive than inline ads. Users seem to prefer not to have to divert their attention from their searching task or take additional steps to close the popup or pop-under windows.

  52. 2004-galletta.pdf: ⁠, Dennis F. Galletta, Raymond Henry, Scott McCoy, Peter Polak (2004; advertising⁠, technology):

    Web page loading speed continues to vex users, even as broadband adoption increases. Several studies have addressed delays in the context of Web sites as well as interactive corporate systems, and have recommended a wide range of ‘rules of thumb’. Some studies conclude that response times should be no greater than 2 seconds while other studies caution on delays of 12 seconds or more. One of the strongest conclusions was that complex tasks seemed to allow longer response times. This study examined delay times of 0, 2, 4, 6, 8, 10, and 12 seconds using 196 undergraduate students in an experiment. Randomly assigned a constant delay time, subjects were asked to complete 9 search tasks, exploring a familiar and an unfamiliar site. Plots of the dependent variables performance, attitudes, and behavioral intentions, along those delays, suggested the use of non-linear regression, and the explained variance was in the neighborhood of 2%, 5%, and 7%, respectively. Focusing only on the familiar site, explained in attitudes and behavioral intentions grew to about 16%. A sensitivity analysis implies that decreases in performance and behavioral intentions begin to flatten when the delays extend to 4 seconds or longer, and attitudes flatten when the delays extend to 8 seconds or longer. Future research should include other factors such as expectations, variability, and feedback, and other outcomes such as actual purchasing behavior, to more fully understand the effects of delays in today’s Web environment.

  53. 2006-galletta.pdf: “When the Wait Isn't So Bad: The Interacting Effects of Website Delay, Familiarity, and Breadth”⁠, Dennis F. Galletta, Raymond M. Henry, Scott McCoy, Peter Polak

  54. 2015-hohnhold.pdf: ⁠, Henning Hohnhold, Deirdre O’Brien, Diane Tang (2015; advertising⁠, economics⁠, statistics  /​ ​​ ​decision⁠, technology):

    Over the past 10+ years, online companies large and small have adopted widespread A/​​​​B testing as a robust data-based method for evaluating potential product improvements. In online experimentation, it is straightforward to measure the short-term effect, ie., the impact observed during the experiment. However, the short-term effect is not always predictive of the long-term effect, ie., the final impact once the product has fully launched and users have changed their behavior in response. Thus, the challenge is how to determine the long-term user impact while still being able to make decisions in a timely manner.

    We tackle that challenge in this paper by first developing experiment methodology for quantifying long-term user learning. We then apply this methodology to ads shown on Google search, more specifically, to determine and quantify the drivers of ads blindness and sightedness, the phenomenon of users changing their inherent propensity to click on or interact with ads.

    We use these results to create a model that uses metrics measurable in the short-term to predict the long-term. We learn that user satisfaction is paramount: ads blindness and sightedness are driven by the quality of previously viewed or clicked ads, as measured by both ad relevance and landing page quality. Focusing on user satisfaction both ensures happier users but also makes business sense, as our results illustrate. We describe two major applications of our findings: a conceptual change to our search ads auction that further increased the importance of ads quality, and a 50% reduction of the ad load on Google’s mobile search interface.

    The results presented in this paper are generalizable in two major ways. First, the methodology may be used to quantify user learning effects and to evaluate online experiments in contexts other than ads. Second, the ads blindness/​​​​sightedness results indicate that a focus on user satisfaction could help to reduce the ad load on the internet at large with long-term neutral, or even positive, business impact.

    [Keywords: Controlled experiments; A/​​​​B testing; predictive modeling; overall evaluation criterion]


  56. 2017-shiller.pdf: “Will Ad Blocking Break the Internet?”⁠, Ben Shiller, Joel Waldfogel, Johnny Ryan

  57. 2017-pagefair.pdf: {#linkBibliography-(pagefair)-2017 .docMetadata}, Sean Blanchfield (PageFair) (2017-02; advertising⁠, economics⁠, technology):

    This whitepaper presents the primary findings of new research by Professor Benjamin Shiller (Brandeis University), Professor Joel Waldfogel (University of Minnesota and the National Bureau of Economic Research), and Dr. Johnny Ryan (PageFair).

    Research of 2,574 websites over 3 years reveals that adblock has a hidden cost: it not only reduces small and medium publishers’ revenue, it also reduces their traffic.

    Studying the changing rate of desktop adblock usage and traffic rank from April 2013—June 2016 reveals that adblock usage is undermining many websites’ ability to invest in content. Affected websites then attract fewer visitors, and so their traffic declines. The full paper is available from NBER, the U.S. National Bureau of Economic Research.

    This is the adblock paradox: users may avoid ads in the short term, but ultimately undermine the value they can derive from the web. To reverse this phenomenon, publishers must listen to users’ legitimate grievances about online ads and respond by fixing the problems. Once they have remedied the users’ grievances, publishers can choose to serve their ads using technology that adblock companies cannot tamper with.

  58. 2017-pagefair.pdf#page=3: “The Hidden Cost of Adblock: Adblock's impact on website traffic”⁠, Sean Blanchfield (PageFair)


  60. ⁠, Shunyao Yan, Klaus M. Miller, Bernd Skiera (2020-05-14):

    Many online news publishers finance their websites by displaying ads alongside content. Yet, remarkably little is known about how exposure to such ads impacts users’ news consumption. We examine this question using 3.1 million anonymized browsing sessions from 79,856 users on a news website and the quasi-random variation created by ad blocker adoption. We find that seeing ads has a robust negative effect on the quantity and variety of news consumption: Users who adopt ad blockers subsequently consume 20% more news articles corresponding to 10% more categories. The effect persists over time and is largely driven by consumption of “hard” news. The effect is primarily attributable to a learning mechanism, wherein users gain positive experience with the ad-free site; a cognitive mechanism, wherein ads impede processing of content, also plays a role. Our findings open an important discussion on the suitability of advertising as a monetization model for valuable digital content.


  62. 2021-suarez.pdf: ⁠, David Suárez, Begoña García-Mariñoso (2021-04-01; advertising⁠, economics⁠, technology):

    • E-commerce and online advertisement are growing trends.
    • The overall impact of ad blockers is unclear.
    • Using survey data, the effect of ad blocker use on online purchases is quantified.
    • The analysis reveals a positive effect of ad blocker use on e-commerce.
    • In the light of the results stakeholders should consider if the present online ads formats are the most suitable.

    The use of ad blocking software has risen sharply with online advertising and is recognized as challenging the survival of the ad supported web. However, the effects of ad blocking on consumer behavior have been studied scarcely.

    This paper uses techniques on a longitudinal survey of 4411 Internet users in Spain to show that ad blocking has a causal positive effect on their number of online purchases. This could be attributed to the positive effects of ad blocking, such as a safer and enhanced navigation.

    This striking result reinforces the controversial debate of whether current online ads are too bothersome for consumers.

    [Keywords: Ad blockers, advertising avoidance, e-commerce, propensity score matching]

    …This study employs a rich dataset coming from a longitudinal survey. The source of the data is a survey conducted by the Spanish Markets and Competition Authority on the same sample of interviewees in the 4th quarter of 2017 and in the second quarter of 2018 ([dataset]CNMCData, 2019). The sample was designed to be representative of the population living in private households in Spain. The information was provided by 4411 Internet users ≥16 years old. At the baseline time point (fourth quarter of 2017) these individuals were asked if they regularly used ad blocking tools when navigating the web. Additionally, the survey collected information on their socio-demographic characteristics (age, gender, education level and employment status) and on how they used Internet (frequency of use of online services like: GPS navigation services, instant messaging, mobile gaming, social networks, e-mail and watching videos on the phone). 6 months later (second quarter of 2018), the same individuals were asked how many online purchases they had made during the previous 6 months (these included goods and services purchases, irrespective of the form of payment). Thus, the outcome variable (number of online purchases) occurred later than the collection of the ad blocking information and the rest of variables (our X covariates).

    Analysis N Treated Controls Difference (ATT) 95% LCI 95% UCI p-value
    Unmatched 4411 5.084 2.735 2.348
    PSM—NN 1648 5.084 3.325 1.759 0.994 2.523 <0.001
    PSM—KM 4411 5.084 3.733 1.351 0.658 2.044 <0.001
    Stratification on PS quintiles 4411 5.084 3.686 1.398 0.724 2.072 <0.001
    Stratification on PS deciles 4411 5.084 3.774 1.310 0.626 1.994 <0.001
    PSM—NN after CEM pruning (1) 1160 4.979 3.773 1.206 0.165 2.246 0.023
    PSM—NN after CEM pruning (2) 1622 5.082 3.476 1.605 0.830 2.380 <0.001

    Table 2: Estimated average treatment effects of ad blockers on online shopping (number of purchases in 6 months). [ATT: average treatment effect on the treated. PSM: propensity score matching. NN: nearest neighbor. KM: kernel matching. PS: propensity scores. CEM: coarsened exact matching. LCI: lower ⁠. UCI: upper confidence interval. (1) CEM pruning by using use of Internet apps covariates. (2) CEM pruning by using socio-demographic covariates.]

  63. 2015-pagefair.pdf: ⁠, PageFair (2015; advertising⁠, economics⁠, technology):

    In the third annual ad blocking report, PageFair, with the help of Adobe, provides updated data on the scale and growth of ad blocking software usage and highlights the global and regional economic impact associated with it. Additionally, this report explores the early indications surrounding the impact of ad blocking within the mobile advertising space and how mobile will change the ad blocking landscape.

    Table of Contents: · 1. Introduction · 2. Table of Contents · 3. Key insights · 4. Global ad blocking growth · 5. Usage of ad blocking software in the United States · 6. Usage of ad blocking software in Europe · 7. The cost of blocking ads · 8. Effect of ad blocking by industry · 9. Google Chrome still the main driver of ad block growth · 10. Mobile is yet to be a factor in ad blocking growth · 11. Mobile will facilitate future ad blocking growth · 12. Reasons to start using an ad blocker · 13. Afterword · 14. Background · 15. Methodology · 16. Tables · 17. Tables

    Key Insights: More consumers block ads, continuing the strong growth rates seen during 2013 and 2014. The findings:

    • Globally, the number of people using ad blocking software grew by 41% year over year.
    • 16% of the US online population blocked ads during Q2 2015.
    • Ad block usage in the United States grew 48% during the past year, increasing to 45 million monthly active users (MAUs) during Q2 2015.
    • Ad block usage in Europe grew by 35% during the past year, increasing to 77 million monthly active users during Q2 2015.
    • The estimated loss of global revenue due to blocked advertising during 2015 was $21.8B.
    • With the ability to block ads becoming an option on the new iOS 9, mobile is starting to get into the ad blocking game. Currently Firefox and Chrome lead the mobile space with 93% share of mobile ad blocking.

  65. ⁠, Anna Sołtysik-Piorunkiewicz, Artur Strzelecki, Edyta Abramek (2019-12; advertising⁠, technology):

    The article shows the main factors of adblocking software usage. The study was based on data obtained by a web questionnaire. The research was focused on evaluation of ad blocking (adblock) software usage factors in five categories: (1) gender, age, and education; (2) use of advertising and sources of knowledge about advertising; (3) technical and social reasons for blocking online advertisements; (4) usage of an adblock-wall; and (5) type of online advertisement. An evaluation of adblock usage factors revealed four main technical reasons for adblock usage connected with website technology and web development problems—interruption, amount of ads, speed, and security; and one social reason for adblock usage, namely, the problem of privacy.

    [Keywords: adblock software, web advertisement, website, security, privacy]


  67. 2019-03-18-googlesurveys-adblocker.csv

  68. 2019-03-21-googlesurveys-adblocknonusagereasons.csv

  69. 2019-03-23-googlesurveys-adblocknonusagereasons-2.csv

  70. 2019-04-07-googlesurveys-adblocker-2.csv

  71. 2019-06-10-googlesurveys-addislike.csv

  72. 2019-03-27-googlesurveys-adblocknonusagereasons-3.csv




  76. #stan-issues

  77. “ Google Analytics Semi-Annual Traffic Report (20170101-20171015)”⁠, Gwern Branwen

  78. Traffic

  79. 2018-04-10-tavan-isadblockingtenpercenthigherthancommonlymeasured.html: {#linkBibliography-(contentpass)-2018 .docMetadata}, Christoph Tavan (contentpass) (2018-04-10; advertising⁠, technology):

    A recent study by contentpass indicates that more than 25% of all ad blockers on desktop devices use the EasyPrivacy blocklist and are therefore invisible to common website analytics software…The by far most popular filter list to block ads is the so-called “Easylist”. It is activated by default in popular ad blockers like Adblock Plus, Adblock or uBlock Origin and focuses on blocking ads both on a network—and on a visual level. Even the built-in ad blocker of Google Chrome uses this list.

    While EasyPrivacy users are now “invisible” to our service as well, we recently integrated our solution under the first party domain on a popular German IT news website. As a consequence of this first party integration the statistics about ad blocker usage were sent to a different URL, which was initially not being blocked by EasyPrivacy. It took about two weeks for the EasyPrivacy community to put the statistics URL of the first party domain on a filter list again.

    These two weeks of unfiltered data allow us to get an idea of how many people use an ad blocker with EasyPrivacy activated (be it Adblock Plus/​​​​Adblock where the user manually activated EasyPrivacy or uBlock Origin where EasyPrivacy is activated by default).

    Our data suggests that over 25% of all users with active ad blocking software on desktop devices use EasyPrivacy and are thus invisible to major web analytics software. In this specific case the true ad blocking rate on desktop was 37% while analytics software that is blocked by EasyPrivacy would only report what corresponds to 27% of ad blocking. Or from a different perspective: 10% of the total desktop traffic on this website is not analyzed and counted by common third party analytics software. Historical data from the time where our service was initially added to EasyPrivacy suggests similar proportions on other sites and verticals.

  80. #replication

  81. 1987-rossi

  82. ⁠, Randall A. Lewis, David H. Reiley (2011-06-08; economics⁠, advertising⁠, statistics  /​ ​​ ​decision⁠, statistics  /​ ​​ ​causality):

    We measure the causal effects of online advertising on sales, using a randomized experiment performed in cooperation between Yahoo! and a major retailer. After identifying over one million customers matched in the databases of the retailer and Yahoo!, we randomly assign them to treatment and control groups. We analyze individual-level data on ad exposure and weekly purchases at this retailer, both online and in stores. We find statistically-significant and economically substantial impacts of the advertising on sales. The treatment effect persists for weeks after the end of an advertising campaign, and the total effect on revenues is estimated to be more than seven times the retailer’s expenditure on advertising during the study. Additional results explore differences in the number of advertising impressions delivered to each individual, online and offline sales, and the effects of advertising on those who click the ads versus those who merely view them. Power calculations show that, due to the high variance of sales, our large number of observations brings us just to the frontier of being able to measure economically substantial effects of advertising. We also demonstrate that without an experiment, using industry-standard methods based on endogenous crosssectional variation in advertising exposure, we would have obtained a wildly inaccurate estimate of advertising effectiveness.

  83. ⁠, Randall A. Lewis, Justin M. Rao (2013-04-23; economics⁠, advertising⁠, statistics  /​ ​​ ​decision):

    Classical theories of the firm assume access to reliable signals to measure the causal impact of choice variables on profit. For advertising expenditure we show, using 25 online field experiments (representing $3.50$2.82013 million) with major U.S. retailers and brokerages, that this assumption typically does not hold. Statistical evidence from the randomized trials is very weak because individual-level sales are incredibly volatile relative to the per capita cost of a campaign—a “small” impact on a noisy dependent variable can generate positive returns. A concise statistical argument shows that the required sample size for an experiment to generate sufficiently informative confidence intervals is typically in excess of ten million person-weeks. This also implies that heterogeneity bias (or model misspecification) unaccounted for by observational methods only needs to explain a tiny fraction of the variation in sales to severely bias estimates. The weak informational feedback means most firms cannot even approach profit maximization.

  84. 2015-lewis.pdf: ⁠, Randall A. Lewis, Justin M. Rao (2015-07-06; advertising⁠, economics⁠, statistics  /​ ​​ ​decision⁠, statistics  /​ ​​ ​causality):

    25 large field experiments with major U.S. retailers and brokerages, most reaching millions of customers and collectively representing $2.8 million in digital advertising expenditure, reveal that measuring the returns to advertising is difficult. The median confidence interval on return on investment is over 100 percentage points wide. Detailed sales data show that relative to the per capita cost of the advertising, individual-level sales are very volatile; a coefficient of variation of 10 is common. Hence, informative advertising experiments can easily require more than 10 million person-weeks, making experiments costly and potentially infeasible for many firms. Despite these unfavorable economics, randomized control trials represent progress by injecting new, unbiased information into the market. The inference challenges revealed in the field experiments also show that selection bias, due to the targeted nature of advertising, is a crippling concern for widely employed observational methods.

  85. Replication

  86. 2019-shapiro.pdf: ⁠, Bradley Shapiro, Günter J. Hitsch, Anna Tuchman (2019-06-11; advertising⁠, economics⁠, statistics  /​ ​​ ​decision⁠, statistics  /​ ​​ ​bias):

    We provide generalizable and robust results on the causal sales effect of TV advertising based on the distribution of advertising elasticities for a large number of products (brands) in many categories. Such generalizable results provide a prior distribution that can improve the advertising decisions made by firms and the analysis and recommendations of anti-trust and public policy makers. A single case study cannot provide generalizable results, and hence the marketing literature provides several meta-analyses based on published case studies of advertising effects. However, results if the research or review process systematically rejects estimates of small, statistically insignificant, or “unexpected” advertising elasticities. Consequently, if there is publication bias, the results of a will not reflect the true population distribution of advertising effects.

    To provide generalizable results, we base our analysis on a large number of products and clearly lay out the research protocol used to select the products. We characterize the distribution of all estimates, irrespective of sign, size, or statistical-significance. To ensure generalizability we document the robustness of the estimates. First, we examine the sensitivity of the results to the approach and assumptions made when constructing the data used in estimation from the raw sources. Second, as we aim to provide causal estimates, we document if the estimated effects are sensitive to the identification strategies that we use to claim causality based on observational data. Our results reveal substantially smaller effects of own-advertising compared to the results documented in the extant literature, as well as a sizable percentage of statistically insignificant or negative estimates. If we only select products with statistically-significant and positive estimates, the mean or median of the advertising effect distribution increases by a factor of about five.

    The results are robust to various identifying assumptions, and are consistent with both publication bias and bias due to non-robust identification strategies to obtain causal estimates in the literature.

    [Keywords: advertising, publication bias, generalizability]

  87. 2021-shapiro.pdf: ⁠, Bradley T. Shapiro, Gunter J. Hitsch, Anna E. Tuchman (2021-07-26; advertising⁠, economics⁠, statistics  /​ ​​ ​decision⁠, statistics  /​ ​​ ​bias):

    We estimate the distribution of television advertising elasticities and the distribution of the advertising return on investment (ROI) for a large number of products in many categories…We construct a data set by merging market (DMA) level TV advertising data with retail sales and price data at the brand level…Our identification strategy is based on the institutions of the ad buying process.

    Our results reveal substantially smaller advertising elasticities compared to the results documented in the literature, as well as a sizable percentage of statistically insignificant or negative estimates. The results are robust to functional form assumptions and are not driven by insufficient or measurement error.

    The ROI analysis shows negative ROIs at the margin for more than 80% of brands, implying over-investment in advertising by most firms. Further, the overall ROI of the observed advertising schedule is only positive for one third of all brands.

    [Keywords: advertising, return on investment, empirical generalizations, agency issues, consumer packaged goods, media markets]

    …We find that the mean and median of the distribution of estimated long-run own-advertising elasticities are 0.023 and 0.014, respectively, and 2 thirds of the elasticity estimates are not statistically different from zero. These magnitudes are considerably smaller than the results in the extant literature. The results are robust to controls for own and competitor prices and feature and display advertising, and the advertising effect distributions are similar whether a carryover parameter is assumed or estimated. The estimates are also robust if we allow for a flexible functional form for the advertising effect, and they do not appear to be driven by ⁠. As we are not able to include all sensitivity checks in the paper, we created an interactive web application that allows the reader to explore all model specifications. The web application is available⁠.

    …First, the advertising elasticity estimates in the baseline specification are small. The median elasticity is 0.0140, and the mean is 0.0233. These averages are substantially smaller than the average elasticities reported in extant meta-analyses of published case studies (Assmus, Farley, and Lehmann (1984b), Sethuraman, Tellis, and Briesch (2011)). Second, 2 thirds of the estimates are not statistically distinguishable from zero. We show in Figure 2 that the most precise estimates are those closest to the mean and the least precise estimates are in the extremes.

    Figure 2: Advertising effects and confidence intervals using baseline strategy. Note: Brands are arranged on the horizontal axis in increasing order of their estimated ad effects. For each brand, a dot plots the point estimate of the ad effect and a vertical bar represents the 95% confidence interval. Results are from the baseline strategy model with δ = 0.9 (equation (1)).

    6.1 Average ROI of Advertising in a Given Week:

    In the first policy experiment, we measure the ROI of the observed advertising levels (in all DMAs) in a given week t relative to not advertising in week t. For each brand, we compute the corresponding ROI for all weeks with positive advertising, and then average the ROIs across all weeks to compute the average ROI of weekly advertising. This metric reveals if, on the margin, firms choose the (approximately) correct advertising level or could increase profits by either increasing or decreasing advertising.

    We provide key summary statistics in the top panel of Table III, and we show the distribution of the predicted ROIs in Figure 3(a). The average ROI of weekly advertising is negative for most brands over the whole range of assumed manufacturer margins. At a 30% margin, the median ROI is −88.15%, and only 12% of brands have positive ROI. Further, for only 3% of brands the ROI is positive and statistically different from zero, whereas for 68% of brands the ROI is negative and statistically different from zero.

    Figure 3: Predicted ROIs. Note: Panel (a) provides the distribution of the estimated ROI of weekly advertising and panel (b) provides the distribution of the overall ROI of the observed advertising schedule. Each is provided for 3 margin factors, m = 0.2, m = 0.3, and m = 0.4. The median is denoted by a solid vertical line and zero is denoted with a vertical dashed line. Gray indicates brands with negative ROI that is statistically different from zero. Red indicates brands with positive ROI that is statistically different from zero. Blue indicates brands with ROI not statistically different from zero.

    These results provide strong evidence for over-investment in advertising at the margin. [In Appendix C.3, we assess how much larger the TV advertising effects would need to be for the observed level of weekly advertising to be profitable. For the median brand with a positive estimated ad elasticity, the advertising effect would have to be 5.37 times larger for the observed level of weekly advertising to yield a positive ROI (assuming a 30% margin).]

    6.2 Overall ROI of the Observed Advertising Schedule: In the second policy experiment, we investigate if firms are better off when advertising at the observed levels versus not advertising at all. Hence, we calculate the ROI of the observed advertising schedule relative to a counterfactual baseline with zero advertising in all periods.

    We present the results in the bottom panel of Table III and in Figure 3(b). At a 30% margin, the median ROI is −57.34%, and 34% of brands have a positive return from the observed advertising schedule versus not advertising at all. Whereas 12% of brands only have positive and 30% of brands only negative values in their confidence intervals, there is more uncertainty about the sign of the ROI for the remaining 58% of brands. This evidence leaves open the possibility that advertising may be valuable for a substantial number of brands, especially if they reduce advertising on the margin.

    …Our results have important positive and normative implications. Why do firms spend billions of dollars on TV advertising each year if the return is negative? There are several possible explanations. First, agency issues, in particular career concerns, may lead managers (or consultants) to overstate the effectiveness of advertising if they expect to lose their jobs if their advertising campaigns are revealed to be unprofitable. Second, an incorrect prior (i.e., conventional wisdom that advertising is typically effective) may lead a decision maker to rationally shrink the estimated advertising effect from their data to an incorrect, inflated prior mean. These proposed explanations are not mutually exclusive. In particular, agency issues may be exacerbated if the general effectiveness of advertising or a specific advertising effect estimate is overstated. [Another explanation is that many brands have objectives for advertising other than stimulating sales. This is a nonstandard objective in economic analysis, but nonetheless, we cannot rule it out.] While we cannot conclusively point to these explanations as the source of the documented over-investment in advertising, our discussions with managers and industry insiders suggest that these may be contributing factors.


  89. 2011-lewis.pdf: ⁠, Randall A. Lewis, Justin M. Rao, David H. Reiley (2011-03; advertising⁠, economics⁠, statistics  /​ ​​ ​decision⁠, statistics  /​ ​​ ​causality):

    Measuring the causal effects of online advertising (adfx) on user behavior is important to the health of the WWW publishing industry. In this paper, using three controlled experiments, we show that observational data frequently lead to incorrect estimates of adfx. The reason, which we label “activity bias”, comes from the surprising amount of time-based correlation between the myriad activities that users undertake online.

    In Experiment 1, users who are exposed to an ad on a given day are much more likely to engage in brand-relevant search queries as compared to their recent history for reasons that had nothing do with the advertisement. In Experiment 2, we show that activity bias occurs for page views across diverse websites. In Experiment 3, we track account sign-ups at a competitor’s (of the advertiser) website and find that many more people sign-up on the day they saw an advertisement than on other days, but that the true “competitive effect” was minimal.

    In all three experiments, exposure to a campaign signals doing “more of everything” in given period of time, making it difficult to find a suitable “matched control” using prior behavior. In such cases, the “match” is fundamentally different from the exposed group, and we show how and why observational methods lead to a massive overestimate of adfx in such circumstances.

    [Keywords: advertising effectiveness, browsing behavior, causal inference, field experiments, selection bias]







  96. 2019-gordon.pdf: ⁠, Brett R. Gordon, Florian Zettelmeyer, Neha Bhargava, Dan Chapsky (2019-05-04; statistics  /​ ​​ ​causality⁠, advertising):

    Measuring the causal effects of digital advertising remains challenging despite the availability of granular data. Unobservable factors make exposure endogenous, and advertising’s effect on outcomes tends to be small. In principle, these concerns could be addressed using randomized controlled trials (RCTs). In practice, few online ad campaigns rely on RCTs and instead use observational methods to estimate ad effects. We assess empirically whether the variation in data typically available in the advertising industry enables observational methods to recover the causal effects of online advertising. Using data from 15 U.S. advertising experiments at Facebook comprising 500 million user-experiment observations and 1.6 billion ad impressions, we contrast the experimental results to those obtained from multiple observational models. The observational methods often fail to produce the same effects as the randomized experiments, even after conditioning on extensive demographic and behavioral variables. In our setting, advances in causal inference methods do not allow us to isolate the exogenous variation needed to estimate the treatment effects. We also characterize the incremental explanatory power our data would require to enable observational methods to successfully measure advertising effects. Our findings suggest that commonly used observational approaches based on the data usually available in the industry often fail to accurately measure the true effect of advertising.

  97. ⁠, Dean Eckles, Eytan Bakshy (2017-06-14):

    Peer effects, in which the behavior of an individual is affected by the behavior of their peers, are posited by multiple theories in the social sciences. Other processes can also produce behaviors that are correlated in networks and groups, thereby generating debate about the credibility of observational (i.e. nonexperimental) studies of peer effects. Randomized field experiments that identify peer effects, however, are often expensive or infeasible. Thus, many studies of peer effects use observational data, and prior evaluations of causal inference methods for adjusting observational data to estimate peer effects have lacked an experimental “gold standard” for comparison. Here we show, in the context of information and media diffusion on Facebook, that high-dimensional adjustment of a nonexperimental control group (677 million observations) using propensity score models produces estimates of peer effects statistically indistinguishable from those from using a large randomized experiment (220 million observations). Naive observational estimators overstate peer effects by 320% and commonly used variables (e.g., demographics) offer little bias reduction, but adjusting for a measure of prior behaviors closely related to the focal behavior reduces bias by 91%. High-dimensional models adjusting for over 3,700 past behaviors provide additional bias reduction, such that the full model reduces bias by over 97%. This experimental evaluation demonstrates that detailed records of individuals’ past behavior can improve studies of social influence, information diffusion, and imitation; these results are encouraging for the credibility of some studies but also cautionary for studies of rare or new behaviors. More generally, these results show how large, high-dimensional data sets and statistical learning techniques can be used to improve causal inference in the behavioral sciences.


  99. ⁠, Alexander Coppock, Seth J. Hill, Lynn Vavreck (2020-09-02; sociology⁠, advertising):

    Evidence across social science indicates that average effects of persuasive messages are small. One commonly offered explanation for these small effects is heterogeneity: Persuasion may only work well in specific circumstances. To evaluate heterogeneity, we repeated an experiment weekly in real time using 2016 U.S. presidential election campaign advertisements. We tested 49 political advertisements in 59 unique experiments on 34,000 people. We investigate heterogeneous effects by sender (candidates or groups), receiver (subject partisanship), content (attack or promotional), and context (battleground versus non-battleground, primary versus general election, and early versus late). We find small average effects on candidate favorability and vote. These small effects, however, do not mask substantial heterogeneity even where theory from political science suggests that we should find it. During the primary and general election, in battleground states, for Democrats, Republicans, and Independents, effects are similarly small. Heterogeneity with large offsetting effects is not the source of small average effects.


  101. 2003-ansolabehere.pdf: ⁠, Stephen Ansolabehere, John M. de Figueiredo, James M. Snyder Jr. (2003-12-01; economics):

    Two extreme views bracket the range of thinking about the amount of money in U.S. political campaigns. At one extreme is the theory that contributors wield considerable influence over legislators. Even modest contributions may be cause for concern and regulation, given the extremely large costs and benefits that are levied and granted by government. An alternative view holds that contributors gain relatively little political leverage from their donations, since the links from an individual campaign contribution to the election prospects of candidates and to the decisions of an individual legislators are not very firm. Although these theories have different implications, they share a common perspective that campaign contributions should be considered as investments in a political marketplace, where a return on that investment is expected.

    In this paper, we begin by offering an overview of the sources and amounts of campaign contributions in the U.S. In the light of these facts, we explore the assumption that the amount of money in U.S. campaigns mainly reflects political investment. We then offer our perspective that campaign contributions should be viewed primarily as a type of consumption good, rather than as a market for buying political benefits. Although this perspective helps to explain the levels of campaign contributions by individuals and organizations, it opens up new research questions of its own.




  105. Faces

  106. GPT-2


  108. ads-triplebyte-banner.png

  109. RNN-metadata

  110. TWDNE

  111. Red





  116. ⁠, Susan Athey, Markus Mobius, Jeno Pal (2021-04):

    A policy debate centers around the question how news aggregators such as affect traffic to online news sites. Many publishers view aggregators as substitutes for traditional news consumption while aggregators view themselves as complements because they make news discovery easier.

    We use Spain as a because in response to a copyright reform enacted in December 2014. We compare the news consumption of a large number of Google News users with a synthetic control group of similar non-Google News users. We find that the shutdown of Google News reduces overall news consumption by about 20% for treatment users, and reduces page views on publishers other than Google News by 10%. This decrease is concentrated around small publishers. We further find that users are able to replace some but not all of the types of news they previously read. Post-shutdown, they read less breaking news, hard news, and news that is not well covered on their favorite news publishers.

    These news categories explain most of the overall reduction in news consumption, and shed light on the mechanisms through which aggregators interact with traditional publishers.