The Retail category in the United States has been among the hardest hit during COVID-19. This stands to reason as non-essential businesses were forced to close and consumers sheltered at home. In this report, we look at how in-store shopping behavior has been disrupted over the course of 2020.
Flurry Analytics, owned by Verizon Media, is used in over 1 million mobile applications, providing aggregated insights across more than 2 billion mobile devices per month. For this analysis, Flurry hand-curated a sample of apps used by vendors to process in-store transactions. Since these apps are primarily launched only when transactions are executed, we used app sessions as a proxy for shopping transactions. Let’s start by looking at how the volume of in-store payments compares to U.S. retail sales.
In the chart above, we compare monthly in-store payment app sessions in blue against retail store sales in grey, from January through July. Retail store sales are seasonally adjusted by the U.S. Census Bureau and cover a broad range of businesses such as clothing, food, electronics and automotive. The first thing you’ll notice is that in-store payment sessions and retail store sales are strongly related, with a correlation coefficient of 0.7. The more frequently consumers make in-store purchases in our sample of apps, the more sales are generated by retail stores. Among other things, this gives us confidence that we are using a relevant sample from Flurry’s data set. When it comes to a shift in shopping behavior, both the volume and amount in dollars of retail transactions began to decline in March, followed by a strong dip in April, and then a steady upward climb from May to July. While we believe that in-store payments are still below what they would have been had COVID-19 never occurred, the rebound has been solid. Retail sales were most affected from March to April, declining by 13%. This decline of in-store sales is the biggest drop observed by the Department of Commerce since record-keeping started in 1992. To identify more precisely when shopping behavior changed, let’s drill down to a daily view of in-store payments.
The chart above shows that from March 16 to early April, as states began to issue shelter-in-place orders and close non-essential businesses, in-store payment app sessions declined by 26%. Over the next 4 months, in-store payment app sessions steadily climbed back by 43%. As some retail businesses gradually re-opened with curbside pickup, and many states lifted stay-at-home orders, in-store shopping returned to pre-pandemic levels. One interesting thing that stands out is how the shape of the data changed before and after shelter-in-place orders were issued. Before, you’ll notice many more blue spikes which denote peak days in payments app usage. After, you’ll see more white dips (and fewer blue spikes) which means payment app usage dropped heavily on many more days. The next chart helps show what’s happening; namely that consumers used to make high volume of payments on the weekend (blue spikes), while once shelter-in-place orders were issued, consumers began to reduce weekend payments (white dips). Let’s take a look.
The chart above shows the average percent change of in-store payment app sessions for each day of the week. We averaged sessions from January 1 to March 15 (pre-shelter-in-place orders) and from March 16 to August 16 (post-shelter-in-place orders). We chose March 16 as the cut-off date since that’s when the majority of shelter-in-place closures took effect in the U.S. and when in-store consumer transactions began to change. Beginning in mid-March, in-store transactions declined across all days of the week by an average of 10%. The likely cause is that consumers are making fewer in-store transactions in order to avoid in-person contact, which increases the risk of exposure to the virus. Instead, a lot of shopping has shifted online. Additionally, what stands out is that in-store transactions fell the most on weekends, by an average of 19%, compared to 10% on weekdays. We believe that working from home has brought about a new weekly schedule where consumers have more time on weekdays to step out and run errands. As a result, we think consumers began to spread their shopping activities more evenly across the week, to open up more time on weekends for leisure activities, and to avoid weekend crowds.
While in-store shopping transactions and retail sales have recovered substantially since the pandemic first struck, consumers unusually continue to shop less on the weekend than during the week. The long shutdown continues to have a lasting impact on the shopping landscape. With the back-to-school shopping season underway, we’ll continue to monitor trends in shopping behavior at retail stores. To get and share our latest mobile insights, follow us on Twitter and LinkedIn.