- We have moved to @Adaptv and @AOLPlatforms. Follow us there for the latest on targeted TV advertising with ROI measurement!
Study Finds TV Advertisers Produce Higher Lift When Using Set Top Box Data Compared to Age/Gender Data Sets
In a peer-reviewed scientific paper accepted by the IEEE (Institute of Electrical and Electronics Engineers) International Conference on Data Mining, “A High-Dimensional Set Top Box Ad Targeting Algorithm including Experimental Comparisons to Traditional TV Algorithms,” three data scientists from the TV targeting company PrecisionDemand have found that targeting consumers with advanced data sets, including set top box data, generates significant sales lift and less wasted spending for TV advertisers, compared to “traditional” age and gender targeting.
Traditionally, TV advertisers have used simple age and gender demographic information to buy media and target their viewers. The study found that while this practice of “audience” targeting can help advertisers reach their potential buyers and scores positively in terms of relevance, set top box targeting, or “buyer” targeting, delivers significantly more buyers per million viewers. The results show that set top box targeting is significantly stronger for TV advertisers as it is based on a broader set of variables.
Targeting “buyers” is a very new concept in TV. To date, TV has been a broadcast, mass-market medium, but it is changing to a precision-targeted medium. The proliferation of networks and the bifurcation of audience has made it necessary to micro-target commercials to relevant programs. Yet previous attempts to measure TV advertising effectiveness have missed the concept of targeting. There is an assumption that advertisers know exactly who their target consumers are, but that hasn’t been true.
“Buyer targeting” is the practice of mining set top box data, as well as advertiser first-party customer data, to build demographic profiles and then look for TV programs that have the highest match to the customer profile. This profile is the advertiser’s true audience target, or “buyer” – the consumer most likely to purchase the good or service.
The researchers tested this targeting on four live television campaigns comprising over 22,000 airings. The experiments were conducted on advertisers that leveraged PrecisionDemand’s DemandFinder technology, which identifies the programs that attract the most “buyers” per 1 million viewers.
- “Audience” targeting and “buyer” targeting are not the same. Audience targeting represents the traditional method of buying media based on age and gender and can yield positive results and reach relevant consumers. “Buyer” targeting, the practice of mining set top box data and the advertiser’s customer data, delivers higher results – lifts in sales, inbound qualified customers, etc. – than audience targeting in every instance.
- Age & gender-based buying (audience targeting) comes with a tremendous amount of waste. Truly targeted TV advertising delivers a greater number of qualified consumers (“buyers”) per million viewers, reducing the amount of waste.
- Set top box targeting, or buyer targeting, does not eliminate age and gender completely. Rather, this type of targeting comprises a superset of available variables including consumer age, gender, and other audience attributes.
- Of note, in one experiment, the team found that buying media based on age and gender information performed worse than if the media were purchased randomly, without any kind of targeting. This particular advertiser was looking for wealthy senior citizens; using age & gender targeting, they were only able to target consumers older than 65 years old, a demographic that is dominated by low-income consumers.
About the IEEE acceptance process:
IEEE is the world's largest professional association dedicated to advancing technological innovation and excellence for the benefit of humanity. IEEE and its members inspire a global community through IEEE's highly cited publications, conferences, technology standards, and professional and educational activities. IEEE, pronounced "Eye-triple-E," stands for the Institute of Electrical and Electronics Engineers.
Papers undergo a double-blind peer review process, and the acceptance rate for full papers was 11.6%. Brendan Kitts and team will present the findings at the IEEE International Conference on Data Mining in Dallas on Dec. 7.
Brendan Kitts, Chief Scientist, PrecisionDemand
Dyng Au, Senior Data Engineer, PrecisionDemand
Brian Burdick, Chief Technology Officer, PrecisionDemand
PrecisionDemand’s marketing platform utilizes big data to provide insights and drive effectiveness and compelling efficiencies for television advertisers. We possess strategic technological capabilities that far exceed those otherwise currently available -- using proprietary technology and better data to enable us to append marketers’ data to millions of TV viewers. We’re effective because we can apply our platform’s technology to realize our clients’ full opportunity on TV and then predict and measure the media’s efficiency with far greater precision than ever before. Headquartered in Seattle and New York, the PrecisionDemand team is made up of industry leading experts in media, technology, and data analysis.
WIT Strategy, for PrecisionDemand
Rich Cherecwich, 774.254.0952