Research Projects

Advertising in the Long Tail: Effects of Website Quality on Advertising Effectiveness

Because of the price difference for short and long tail websites, the decision whether to advertise in the long tail has substantial managerial relevance. However, companies that decide to explore the long tail face the challenge that their ads may appear on websites with low quality (i.e., non-premium websites). The recent example of companies abandoning their ads on Google because their ads appeared on websites with inappropriate content underlines the fact that website quality is difficult to be controlled in the long tail, and that managers perceive website quality as an important determinant of online advertising effectiveness. Still, despite the high managerial relevance, little is known about how website quality affects online advertising effectiveness. This research addresses this gap, investigates the effect of website quality on online advertising effectiveness, and provides managerial implications on whether, and for which types of brands exploiting the long tail may be a risky strategy.

Get in touch with Nadia Abou Nabout  to learn more about the project!

Disclosing information about ad placements in online display advertising

The effect on demand and price Real-time bidding is an auction-based ad selling mechanism in real-time and considered to be a revolution for online display advertising. Despite the ability to bid for single targeted impressions through ad exchanges, advertisers very often do not know on which websites their ads appear. Hence, they frequently set up their campaigns to run on a bundle of websites. In practice, ad exchanges could easily grant website information to advertisers in bid requests, which, in turn, allows them to choose individual websites to bid for. However, ad exchanges are reluctant to disclose information about where the ad is run (i.e., the ad placement) as advertisers could start cherry-picking websites, such that ad inventory on some websites might remain unserved and overall price levels could decrease.

The aim of this project, thus, is to investigate the effect of granted ad placement information on demand as well as on price by taking advantage of the recent information disclosure about ad placements by a major European ad exchange. Moreover, following industry categorizations, we distinguish between premium and non-premium website quality to better understand whether advertisers start to cherry-pick premium websites. This project provides insights for ad exchanges selling advertising inventory into information disclosure decisions by examining advertisers’ reactions to the granted information in terms of demand and price and quantifying the economic impact of information disclosure in the context of online display advertising, particularly in real-time bidding.

Get in touch with Sila Ada to learn more about the project!

The Impact of Display Ad Viewability on Advertising Effectiveness

Viewability, i.e., pixel percentage in view and exposure time of an online display ad, is a recently available feature in advertisers’ campaign dashboards. This project evaluates the impact of ad viewability on advertising effectiveness and, therefore, provides guidance for practitioners to target the right viewability values for their campaigns goals. We measure advertising effectiveness in terms of branding outcomes (brand awareness and recall) as well as behavioral consumer responses (clicks, visits to a website, and conversions). Consequently, we are able to provide insights into whether ad viewability needs to be higher for branding or behavioral outcomes to be achieved.

Get in touch with Christina Uhl to learn more about the project!

Do Publishers Miss out on Higher Revenue by Cluttering their Websites?

According to eMarketer (2016), more than half (54%) of ad buyers state that “ad clutter” (i.e., the number of ad slots on a webpage) is the biggest obstacle of digital advertising growth in the US market. Since publishers’ (i.e., website owners and, thereby, content creators) revenue is directly linked to the number of ad slots on their webpage, they have a clear incentive to increase the number of ad slots on a page. However, if ads on a webpage start to take over too much space, webpage visitors might get overwhelmed. Hence, visitors might start to ignore ads leading to more banner blindness and a substantial decrease in banner ad performance (e.g., click-through rate (CTR)). Moreover, the amount of “ad clutter”, which is being tracked by brand safety and ad verification programs, is recently put to use by advertisers when making decisions on where to run their ads. Consequently, publishers might risk missing out on higher revenue by cluttering their webpages.

The aim of this project is to evaluate the impact of ad clutter (i.e., the number of ad slots on a webpage) on ad performance (i.e., click-through rate) as well as on the revenue generated by a specific website (ad impressions sold on website x price per impression). In addition, we examine how the user’s goal (i.e., being goal-oriented versus browsing-oriented while surfing the Internet) moderates this relationship. This project offers important implications for both, advertisers and publishers: For advertisers, it provides recommendations on which websites to advertise on and how much to bid given the website’s level of ad clutter. For publishers, it provides insights into how their revenue is affected (in the short- and possibly the long-term) by a change in the number of ad slots on the website.

Get in touch with Sila Ada to learn more about the project!

Online Content Monetization: Paywall Introduction and the Effect on Influencer Success

Many online platforms (e.g., video streaming platforms such as YouTube) allow influencers such as PewDiePie to monetize their content through advertising, or subscription fees, or both. Influencers therefore face the challenge of finding the right mix of income streams. While relying on advertising income and providing content free of charge might increase influencer popularity by attracting a larger audience, putting content behind a paywall and charging for it will likely increase the per-view income from highly-engaged consumers that are willing to pay for exclusive content. However, this strategy entails the risk of decreasing the influencer’s popularity by restricting access to their content. In addition, charging for content that was previously available for free might lead to a consumer backlash, which might, in turn, negatively affect influencer success. So far, no conclusive evidence exists regarding the question, which of these effects dominates.

To understand how these opposing effects play out and how the introduction of a paywall affects influencer success, we exploit a natural experiment that recently occurred when the largest video sharing platform worldwide introduced a paywall for exclusive content (i.e., YouTube Red), which allowed content providers to monetize content through subscription fees in addition to generating income through advertising. This unique setting allows us to analyze how the introduction of a paywall affects influencer popularity and online chatter. Specifically, we investigate (1) how the paywall introduction affects influencer popularity (i.e., video views and channel subscribers) and online chatter (e.g., shares, likes on other social media platforms), and (2) how it moderates the relationship between online chatter and influencer popularity. To test these effects, we analyze a dataset comprising more than 1,500 influencers and more than 162 billion video views over the period of 16 months. We compare the key indicators (e.g., subscribers, plays likes, shares) before and after the paywall launch and relative to those influencers that did not adopt a paywall for their content. Influencers might use these insights to optimize their mix of income streams.

Get in touch with Christian Hotz-Behofsits to learn more about the project!

The Reduced Customer Revenue of TV Induced Online Shoppers

Second screen usage has dramatically increased within the last years and many consumers spontaneously decide to visit an online shop in the few minutes after watching the shop’s TV ad. In two large-scale empirical studies, we examine the behavior of these TV-induced customers along the sales funnel (including conversions, shopping baskets, and repeat purchase behavior). Our studies are based on two unique data sets provided by European online retailers that both run extensive TV advertising. Since consumers cannot be observed in their living-room, we develop a Bayesian model to determine the probability that a given visit to the online shop is a spontaneous reaction to a TV ad. We then model differences in consumer behavior between TV-induced customers and their non-induced counterparts.

Get in touch with Christina Uhl to learn more about the project!