Marketing Mix Modeling of Traffic to the Store Under the Covid-19 Crisis

Abstract

The paper contains the results of marketing mix modeling for Ukrainian retail in conditions of the COVID-19 crisis. The main goals of the research are modeling the level of traffic to the store based on regression analysis and forming appropriate recommendations for media strategy. Estimating the influence of media on business KPI makes a basis for ROI calculations and optimization of budget allocation between communication channels by periods, formats, and optimization of media pressure. Models for offline and online traffic were constructed based on weekly data for 2018-2021 and since 2020 there is a strong impact of COVID-19 on traffic and media response. In 2020 there was a significant drop in offline traffic due to the lockdown, but also there was deferred demand, which was compensating for a part of the traffic. The results show that TV is the main driver for offline traffic and digital - for online, but there are also significant impacts of TV and digital on online and offline traffic, respectively. During the lockdown, the mobility of consumers dropped, that is, a decrease in response from Out of Home advertising; therefore, we need to compensate for this by higher activity in other media channels. Scenario forecasting of different media mix helps to select the most efficient strategy taking into account memory decay of advertising, period of activity, and weekly weights. Marketing mix modeling is an effective tool for business management, as it generates opportunities to improve ROI by more than 15% and ensures the achievement of business goals in the most efficient way.
 Keywords: marketing mix modeling, COVID-19, ROI, regression analysis, retail, traffic