Media Multiplexing Behavior: Implications for Targeting and Media Planning

TLDR

Consumers increasingly consume fragmented “chunks” of multiple media types—TV, radio, Internet, and print—in rapid succession. The study aims to forecast media consumption by modeling media multiplexing, predicting which media or combinations audiences will use, and understanding substitution and complementarity effects. The authors develop a forecasting model that integrates media‑multiplexing behavior, interdependencies, and consumer heterogeneity, calibrated with individual media diaries, and introduce a utility function that captures cross‑channel complementarities through interactive satiation effects. Accounting for media synergies within a single utility specification markedly improves forecast accuracy, and individual‑level analyses reveal detailed insights into media switching, multiplexing, and heterogeneity that aggregate data miss.

Abstract

There is a growing trend among consumers to serially consume small, incomplete “chunks” of multiple media types—television, radio, Internet, and print—within a short time period. We refer to this behavior as media multiplexing and note that key challenges for integrated marketing communications media planners are (1) predicting which media or combination of media their target audience is likely to consume at any given time and (2) understanding potential substitutions and complementarities in their joint consumption. We propose a forecasting model that incorporates media-multiplexing behavior of both traditional and new media, their interdependencies, and consumer heterogeneity, and we calibrate the model using a rich database of individual-specific media activity diaries. The results suggest that accounting for media synergies within a single utility specification significantly improves model forecasts. We also introduce a utility function that directly models cross-channel media complementarities via interactive effects of the satiation parameters of own and joint consumption of various media types. Finally, our individual-level analyses generate unique insights on consumer-level media switching, multiplexing, and individual heterogeneity often ignored in aggregate data.