A look at the challenges of online grocery shopping in US, where 22% of apparel sales and 30% of electronics sales happen online versus just 3% of grocery sales
Alana Semuels / The Atlantic : Tweets: @alexhern Tweets: Alex Hern / @alexhern : My own theory about why US online grocery shopping sucks is that US *offline* grocery shopping sucks in ways that have to be fixed before consumer behaviour can take to the internet http://www.theatlantic.com/...
Context & Ripple Effects
In 2019, The Atlantic framed online grocery as the outlier of US e-commerce: 22% of apparel and 30% of electronics sales happen online against just 3% of groceries, with Alex Hern arguing the real blocker is that American supermarkets are bad enough in person that shoppers have no digital habit to migrate from. Alana Semuels' reporting on driving for Amazon Flex and clearing less than minimum wage after expenses supplied the cost-side explanation: last-mile grocery delivery is expensive to staff.
Four years on, the numbers barely moved. Even after the pandemic's forced adoption, Numerator put Amazon at roughly 3% US grocery share including Whole Foods, UBS counted about 90% of grocery sales still happening in physical stores in 2022, and overall e-commerce's share of US retail had slipped back toward pre-pandemic levels as in-store buying recovered.
First-order effects
- Amazon's grocery push stalled at single-digit share despite owning Whole Foods and its own delivery workforce, leaving conventional supermarket chains still controlling roughly nine in ten US grocery dollars.
- Delivery economics remain broken at the margin: Semuels' sub-minimum-wage Flex earnings show the per-order labor cost that keeps grocery e-commerce unprofitable where apparel shipping is not.
Second-order effects
- With grocery digitization failing, retailers who bet on online growth face the same reversal documented in UK fashion, where shoppers shifting back to physical stores compounded online players' logistics costs — pushing capital back into store networks.
- Thin-margin grocers avoid subsidizing delivery at scale, which caps volume, which in turn prevents the density needed to lower per-order fulfillment cost — a loop competitors like Amazon have not cracked either.
Third-order effects
- If the pattern holds, grocery becomes the structural anchor of physical retail: a category whose perishability, low margins, and labor-intensive fulfillment resist the digitization curve that transformed apparel and electronics.
- E-commerce penetration looks category-specific rather than inevitable — post-pandemic data showing overall online share retreating suggests adoption follows unit economics and habit, not a one-way ratchet.
The trend: US retail digitization is proving uneven and partially reversible, with grocery remaining the durable physical-retail holdout even after a pandemic-scale adoption shock.