Survey-based measurement of the adoption of grocery delivery services: A commentary

This note tries to identify best practices in the use of consumer surveys as a tool to quantify the uptake of grocery delivery services. The prime motivation is an article by Wang et al. (2021), who examine how the COVID-19 pandemic has affected the use of such services in the United States. I argue that Wang et al. risk underestimating both new adoption and continuance intention because they model these decisions as individual-level choices. In particular, Wang et al.'s sample likely comprises a substantial share of respondents who are not the primary grocery shopper of their household. When confronted with a survey question that inquires whether they themselves have had groceries delivered, these respondents may well answer 'no' - even in situations where the household did, in fact, make use of such a service. In other words, there is a danger of false negatives. I also point out that there are other recent papers which suffer from the same methodological issue, and I indicate ways to avoid it.