Setting a Planning Posture in a Crisis and Post-Crisis Environment

MFA Background Pics light bulb

As a result of the COVID pandemic, consumer behavior has changed across multiple dimensions – where consumers shop, how often, how much is bought and how much is consumed, at home vs out of home, etc. These changes along with the predictions of virus containment will influence categories and brands beyond the heat of the pandemic into the near future. Market Fusion Analytics (MFA) is leveraging its 20+ years of expertise in behavioral sciences to develop short-term and long-term category sales forecasts that account for changes in consumer behavior.

The causal models are built bottom-up by testing a rich set of predictors collected weekly at a granular level.

We combine our internal COVID-19 related projections using pandemic metrics at a granular geography level, data from multiple 3rd parties, and internal sales data to develop business forecasts which help our clients navigating this challenging environment. Our forecasts are derived from dynamic models and based on a set of assumptions that evolve over time as new information becomes available. We create an on-going data feed to track the accuracy of predictions and update projections to ensure better forecasting accuracy.

MFA’s Econometric Modeling isolates and quantifies the individual impact of multiple drivers of revenue sales, which includes the impact of COVID-19.

Our 2-prong modeling approach uniquely links predictions of overtime impact from COVID-19 based on MFA’s “Fear Factor” model and the omnichannel category sales forecasts. As a result, our forecasting models account for both category dynamics in the world of changing consumer behavior and the on-going impact of the COVID-19 pandemic on a specific vertical and sales channel. Unlike black-box models based on extrapolation, MFA’s “Fear Factor” model uses causal drivers such as the socio-political environment, density of population, demographics, social mobility, evolving consumer sentiments, etc., to predict the impact of the pandemic on specific category behavior at a local level. As a result, our data-driven approach provides an explanation (rationalization) for the projections which is key to validation and adoption.

If you are interested in learning more, please contact Tamir Choina at