Real-Time Strategies in Digital Markets
(A2019-9722)

AUTHORS

Siham El Kihal, Frankfurt School of Finance & Management ; Alexander Bleier, Frankfurt School of Finance & Management

ABSTRACT

In today’s highly dynamic digital markets, firms need to continuously adjust their marketing strategies. From four different perspectives, this special session provides insights about how firms can adapt by developing real-time marketing strategies. Using a range of methods and data such as quasi-experiments, social media and transactional records as well as machine learning approaches, each paper contributes novel insights on how managers can improve their real-time strategies in dynamic digital environments.

1. Understanding Real-Time Changes in Viewership during Television Advertisements

As the television industry increasingly moves toward a programmatic ad-buying model, understanding real-time changes in viewership during advertisements is becoming more and more important. The first study by Fossen and Bleier, investigates the relationship between social TV activity, i.e., consumers’ joint viewing of television programming alongside the production or consumption of social media conversations about the program, and viewer retention during ads. The results show how firms can leverage information about online chatter to derive real-time insights for their marketing strategies.

2. Quantifying the Impact of Real-Time B2B Negotiation on B2C platforms

B2C online platforms have grown rapidly over the years. Negotiation between the online platform, as a business entity, and the businesses on the B-side has been widely used as a pricing mechanism. This study by Chu, Zhang, and Manchanda looks at a healthcare online platform, and examines how the negotiation outcomes change as the size of the consumer side or of the business side change. The study shed lights on how a B2C platform can utilize real-time negotiation to adjust to the changing market environment.

3)    Incorporating Images in Real-Time Forecasting of Product Profitability

To make effective merchandising decisions, the third study by Dzyabura, El Kihal, and Ibragimov helps omnichannel retailers in forecasting product return rates prior to their launch, using product image features, extracted via convolutional neural networks.  Retailers can use this novel approach to better forecast a product’s profitability and improve real-time decisions on product merchandising.

4)    Real-Time Marketing’s Role in Affecting Virality and Firm Value

In the fourth study, Borah, Banerjee, Lin, Jain, and Eisingerich, run a quasi-experiment on Twitter, in which they compare a focal brand’s real-time marketing message with a counterfactual synthetic brand message, to provide evidence that the firm’s brand message leads to virality. The study shows when real-time marketing tweets significantly drive firm value and provides direct recommendation to firms on how to manage their communications with consumers on social media platforms.