INFORMS News: Turner Broadcasting receives Innovative Applications in Analytics Award

By Scott Grasman

Turner Broadcasting System, Inc. received the 2017 Innovative Applications in Analytics Award at the 2017 INFORMS Analytics Conference in Las Vegas. The winning entry, “Audience Targeting Solutions Powered by Advanced Analytics,” was presented by Wes Chaar, José Antonio Carbajal and Peter Williams from Turner Broadcasting.

The Innovative Applications in Analytics Award (IAAA), sponsored by Caterpillar and the INFORMS Analytics Society, recognizes creative and unique application of analytical techniques. The prize “promotes the awareness and value of the creative combination of analytics techniques in unusual applications to provide insights and business value.” To win the award, implementations must integrate theoretical advances and innovative applications in order to create value.

Finalists presented their work at the conference to a judging committee chaired by Scott Grasman from Rochester Institute of Technology.

For decades, television advertisement deals had been guaranteed using only primary demographic metrics specified by age and gender. In the last few years, data fusion has allowed viewership data to be fused with frequent shopper card data, credit card data or even custom surveys to construct targeted TV audience segments such as “cereal buyers” or “auto intenders.” These new, more granular audience segments have challenged traditional ways of forecasting, scheduling, and managing inventory in the media industry. Through advanced analytics, Turner has taken the lead in offering targeted ad products that better address the needs of our advertiser and agency partners.

Turner has developed two core audience targeting solutions: TargetingNOW and AudienceNOW. TargetingNOW takes an existing advertising deal, which is still guaranteed on demographic viewership and maintains its original media mix, and optimizes its spot placements to increase the delivery of a secondary targeted segment. AudienceNOW relaxes many, but not all, of the traditional mix constraints to produce fully optimized deals and spot placements that maximize targeted audience delivery across the entire portfolio of Turner’s networks.

A suite of advanced analytics tools power these targeting solutions. On the descriptive side, analysis and visualization tools present detailed historical or predictive information for particular segments and allow comparison of Turner networks against competitors. On the predictive side, a scalable, accurate and data-agnostic forecasting approach called Competitive Audience Estimation can build granular estimates for virtually any audience segment. Finally, on the prescriptive side, large-scale, mixed-integer-programming models optimize deal composition and spot placements.

Second prize was awarded to Valentina Ferretti from London School of Economics and Political Science for her work on “How to Regenerate Disused Railways? An Integrated Analytics Approach.”

Inactive railway lines and disused station buildings constitute an increasingly important heritage asset and are becoming the focus of regeneration processes worldwide. However, the decision of what to do in order to reuse abandoned railways represents a complex decision-making problem, involving heterogeneous impacts and multiple stakeholders leading to conflicting objectives. Such a context calls for the use of analytics able to support transparent, replicable and justifiable processes/results.

This project proposes a combination of different analytics to effectively support collaborative decision-making processes where a decision must be made among competing options. The study developed and tested an integrated analytical approach by combining:

  • preference elicitation analytics with visualization analytics in the descriptive phase of the process;
  • sensitivity analysis with visualization analytics in the predictive phase of the process; and
  • prescriptive decision analysis with facilitated modelling throughout the whole process.

The proposed framework has been tested on a real case study dealing with transportation systems’ planning in Northern Italy, where several passenger railway lines have recently been abandoned and replaced by bus services.

The main objective of the study was to investigate which role integrated decision analytics can play to support heterogeneous impacts’ aggregation in territorial planning, by discussing the operability, applicability and transparency of the developed methodological framework.

The contribution brought by the study is twofold and refers to: (i) improved operability of the proposed tools obtained by combining visualization analytics with consolidated preference elicitation protocols for assessing multiple impacts and (ii) the provision of a replicable working tool for policy-makers.

2018 Award Submissions

Implementations that span applications of descriptive, predictive or prescriptive analytics, as well as data creation, collection and dissemination that support or enable novel analytical methods, can be submitted for the 2018 Innovative Applications in Analytics Award. Applicants must submit a 500-1,000-word summary by Sept. 1 to Committee Chair Juan R. Jaramillo (jaramijr@farmingdale.edu). Semifinalists will be notified by Sept. 30, and finalists will be notified by Dec. 9. Finalist presentations will take place at the INFORMS Analytics Conference in Baltimore, April 15-17, 2018. The winner will present at the INFORMS Annual Meeting in Phoenix, Nov. 4-7, 2018. For more details, contact the committee chair.