During the pandemic, DoorDash glided up to a $72 billion valuation in its December 2020 IPO. The question for investors and companies moving forward is whether and to what extent such changes in consumer behavior will persist, post-pandemic. To answer this, Elliot Oblander and Daniel McCarthy used an event study and a panel regression to quantify the magnitude of the pandemic’s impacts and evaluate the mechanisms driving them.

Analyzing multiple data sources (customer purchase patterns, geographic market coverage for the delivery category, regional economic impact of COVID, and stay-at-home behavior), the researchers charted trends in customer acquisition and retention. They also decomposed the number, frequency and size of purchases. This “bottom-up” decomposition yielded more accurate, forward-looking assessments of overall sales.

In this paper, we quantify the impact of COVID-19 on customer purchase behaviors – customer acquisition, retention, ordering, and spending – within the restaurant food delivery category in the United States and assess the mechanisms through which these effects have arisen using a unique collection of data sources.

Using “gold standard” population-level data and calibrating their model against sales reported by Grubhub and DoorDash, for external validity, the researchers found that category sales would have grown 38% in the absence of the pandemic. That is significantly less than the 122% that actually occurred. Their decomposition further suggests that most, 84%, of this impact, is attributable to existing, pre-COVID customers, rather than to new customers acquired during the pandemic.

This growth among pre-COVID customers is roughly equally due to an increase in purchase frequency and an in average order value (AOV). Since pre-pandemic customer acquisition and purchase trends had been declining, the pandemic’s impact was significant, creating $19.3 billion in incremental sales in 2020, or a 69% overall increase in sales year-on-year.

Read the full working paper here.