TV Viewing and Advertising Targeting

Yiting Deng and Carl F. Mela, 2017, 17-111-05

Digital innovations such as digital video recorders and set-top boxes are both a blessing and a curse for advertisers. On the one hand, set-top boxes have greatly enhanced viewers’ TV-viewing experiences, leading to increased viewing consumption and the opportunity to target advertising at the TV-set level. On the other hand, set-top boxes enable viewers to fast-forward past, and thus avoid, advertisements.

In this report, Yiting Deng and Carl Mela take advantage of micro-level viewing data and micro-targeting capabilities inherent in set-top boxes to better understand viewer behavior and propose ways for advertisers to improve their targeting profitability.

They use a unique dataset that integrates several data sources (set-top box data, purchase data, show data, and advertising data) to develop and estimate models of households’ TV program sampling / consumption and advertising response.

Their viewing model is based on the following observations:

  • Households watch prime-time TV almost every night (average = 85% of days) and for most of the evening (average = 88% of prime-time hours). Thus, any gains from targeting largely arise from what a viewer watches rather than whether they watch.
  • Once a viewing session commences, viewers generally sample several shows prior to selecting one to view. Once a show is selected, viewers typically watch a show to its conclusion. Within a show, viewers’ advertising avoidance is more common when the show is recorded than when it is live.
  • Person-specific factors explain most of the variation in advertising avoidance (20.4%), suggesting the potential for significant returns to household-level advertising targeting. Variations explained by genre, brand, time, and category are respectively 0.7%, 0.3%, 0.0%, and 0.0%, suggesting targeting on these will have smaller effects on advertising avoidance.

Deng and Mela consider several household-level targeting scenarios by manipulating: (1) whether the object is to minimize costs for a given set of views or to maximize incremental profit from advertising, and (2) whether the advertising purchase is made in advance or in real time.

Their results indicate that micro-targeting can lower advertising costs and raise incremental profit even in the face of ad avoidance. For example, with advance buy, it is possible to lower costs per target ad view by over 90%. With real-time buy, it is further possible to lower target costs per view while concurrently increasing target views; in one schedule, views to the target households were increased by 47% while costs concurrently reduced by 7%.

Overall, they find that the greatest potential to increase the profitability of advertising arises from (1) the integration of purchase data with viewing data (single-source data) and (2) the ability to buy placements real-time instead of in advance.

As single-source data becoming increasingly available and with the development of an advertising ecosystem to enable real-time buying on TV, their analysis highlights the importance of these advances and a means of exploiting them. More-effective targeting may open new paths to TV advertising pricing by allowing TV networks and cable companies to sell “boxes” as well as shows to advertisers.

Yiting Deng is an Assistant Professor at the UCL School of Management, University College London. Carl F. Mela is the T. Austin Finch Foundation Professor of Business Administration at the Fuqua School of Business, Duke University and MSI 2017-2019 Executive Director. A portion of this work was completed while the second author was visiting at the Rotterdam School of Management, Erasmus University.

The authors would like to thank Bryan Bollinger, Pedro Gardete, Roni Shachar, Richard Staelin, Peng Sun, Kenneth Wilbur, Yi (Daniel) Xu, participants at the 2016 Marketing Science Conference, and seminar participants at Cornell University, Duke University, Erasmus University, HKUST, McGill University, New York University, Peking University, Syracuse University, University at Buffalo, University of California-Riverside, University College London, University of Miami, University of Notre Dame, and University of Rochester for useful comments. The usual disclaimer applies.

Related links

TV Ads and Search Spikes: Toward a Deeper Understanding
Rex Y. Du, Linli Xu, and Kenneth C. Wilbur (2017) [Report]

Social TV, Advertising, and Sales: Are Social Shows Good for Advertisers?
Beth L. Fossen and David A. Schweidel (2016) [Report]


  • Corporate: FREE
  • Academic: FREE
  • Subscribers: FREE
  • Public: $18.00



Employees of MSI Member Companies enjoy the benefits of complete online access to content, member conferences and networking with the MSI community.



Qualified academics benefit from a relationship with MSI through access to, conferences and research opportunities.



The public is invited to enjoy partial access to content, a free e-newsletter, selected reports and more.



Become a Subscriber

MSI's Online Library of 400+ reports, authored by marketing academics, offers new research and evidence-based insights

Read More

Priority Topics

Stay Informed

The MSI Mailing List

Subscribe to our email list to stay informed about upcoming events, news, etc.

Social Media