Bringing home the bacon: Estimating willingness to pay for autonomous grocery delivery across U.S. households

Abstract

We conduct a discrete choice experiment across 771 U.S. households to estimate willingness to pay (WTP) values for autonomous and human grocery delivery. We estimate that urban households, on average, are willing to pay about $6.70 for same-day human delivery to the household’s doorstep. This value drops by about $1.90 for autonomous delivery, by $2.30 per day of delivery delay, and by $3.10 or $7.90 for delivery to curbside or neighborhood locker locations, respectively. Households located 20 min or more from the nearest major grocery store are willing to pay $3 more for delivery than other households and are more willing to accept delivery delay and curbside or locker delivery locations, while households below the poverty line are less sensitive to delivery delay. Using a latent class logit model, we identify three distinct types of consumers: (1) Delivery Compromisers, representing 45% of weighted respondents, prefer delivery but are sensitive to timing, location, and automation; (2) Delivery Indifferent consumers (39%) have no statistically significant WTP for delivery over in-person shopping; and (3) Delivery Enthusiasts (16%) have high WTP for delivery. Simulation results suggest that if autonomous delivery were able to provide rapid delivery at low cost, it could potentially capture a majority of shopping choices for the Enthusiast groups and a substantial share of choices for Compromisers.

Keywords

• Autonomous vehicles
• Grocery delivery
• Discrete choice