Ghost Ads: A Revolution in Measuring Ad Effectiveness

Johnson, Garrett A.; Lewis, Randall A.; Nubbemeyer, Elmar I

doi:10.2139/ssrn.2620078

Cited by 4 publications

(7 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Running randomized experiments (or field experiments) is becoming the standard approach to measuring the marginal effectiveness of online campaigns (Chittilappilly 2012, Lewis et al 2011, Yildiz and 2 Marketing Science, Articles in Advance, pp. 1-19, © 2016INFORMS Narayanan 2013, Johnson et al 2016. In this practice, the ad is assumed to be the treatment to evaluate, and users are randomly separated into study and control groups.…”

Section: Introduction and Problem Contextmentioning

confidence: 99%

“…Moreover, to display a placebo ad, the opportunity to advertise must be consumed and the campaign must exist in the marketplace (i.e., campaign presence effect). Johnson et al (2016) acknowledge the bias induced by endogenous user selection when running a placebo campaign. Similar to propensity-score based corrections, their proposed solution predicts ad exposures of users in the control group based on their features, which are often noisy and fragmented.…”

Section: Introduction and Problem Contextmentioning

confidence: 99%

“…External validity of effects is implicitly assumed by most evaluation practices in literature(Lewis et al 2011, Yildiz and Narayanan 2013, Johnson et al 2016. Given evolving user tastes, marketplace dynamics, and the sequential learning of targeting algorithms, even medium-term effect generalizations could be highly inaccurate.5 A typical belief is that user pre-treatment feature based balancing is the only way to show that control and study populations are statistically equivalent(Johnson et al 2016). However, Assumptions 1 and 2 do not require these features as long as the user assignment is independent of the effect, i.e., random.…”

mentioning

confidence: 99%

“…Note that the campaign budget, a crucial decision, is captured in the campaign attribution by this estimator. Also, we emphasize that only the visiting users to a given set of publisher websites are potentially selected for ad exposure, and not all of the online users as stated byJohnson et al (2016). Although other causal frameworks have been developed, mainly the Structural Equation framework(Heckman 2008), we use Potential Outcomes to model post-treatment variables with experimental data.7 The current analysis holds for the randomized design inFigure 3(d)if only the control and study arms are analyzed.…”

mentioning

confidence: 99%

“…By focusing on average effect estimates, we obviate the need to predict individual selection for ad exposure indicators, as performed byJohnson et al (2016), with associated prediction errors.…”

mentioning

confidence: 99%

See 4 more Smart Citations

Experimental Designs and Estimation for Online Display Advertising Attribution in Marketplaces

Barajas

Akella

Holtan³

et al. 2016

Marketing Science

View full text Add to dashboard Cite

O nline Display Advertising's importance as a marketing channel is partially due to its ability to attribute conversions to campaigns. Current industry practice to measure ad effectiveness is to run randomized experiments using placebo ads, assuming external validity for future exposures. We identify two different effects, i.e., a strategic effect of the campaign presence in marketplaces, and a selection effect due to user targeting; these are confounded in current practices. We propose two novel randomized designs to: (1) estimate the overall campaign attribution without placebo ads, (2) disaggregate the campaign presence and ad effects. Using the Potential Outcomes Causal Model, we address the selection effect by estimating the probability of selecting influenceable users. We show the ex-ante value of continuing evaluation to enhance the user selection for ad exposure mid-flight. We analyze two performance-based (CPA) and one Cost-Per-Impression (CPM) campaigns with 20 million users each. We estimate a negative CPM campaign presence effect due to cross product spillovers. Experimental evidence suggests that CPA campaigns incentivize selection of converting users regardless of the ad, up to 96% more than CPM campaigns, thus challenging the standard practice of targeting most likely converting users.

show abstract

Section: Introduction and Problem Contextmentioning

confidence: 99%

Section: Introduction and Problem Contextmentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 3 more Smart Citations

Experimental Designs and Estimation for Online Display Advertising Attribution in Marketplaces

Barajas

Akella

Holtan³

et al. 2016

Marketing Science

View full text Add to dashboard Cite

show abstract

Inefficiencies in Digital Advertising Markets

Gordon¹,

Jerath

Katona

et al. 2020

Journal of Marketing

105

View full text Add to dashboard Cite

Digital advertising markets are growing and attracting increased scrutiny. This article explores four market inefficiencies that remain poorly understood: ad effect measurement, frictions between and within advertising channel members, ad blocking, and ad fraud. Although these topics are not unique to digital advertising, each manifests in unique ways in markets for digital ads. The authors identify relevant findings in the academic literature, recent developments in practice, and promising topics for future research.

show abstract

Social Advertising

Tucker

2012

SSRN Journal

View full text Add to dashboard Cite

Ghost Ads: A Revolution in Measuring Ad Effectiveness

Cited by 4 publications

References 32 publications

Experimental Designs and Estimation for Online Display Advertising Attribution in Marketplaces

Experimental Designs and Estimation for Online Display Advertising Attribution in Marketplaces

Inefficiencies in Digital Advertising Markets

Social Advertising

Contact Info

Product

Resources

About