Online shopping generates, on average, about one-third the CO2 of offline shopping, according to an Oliver Wyman study, primarily because it eliminates the car trip. That advantage sounds conclusive. It isn’t. The one-third figure assumes ideal conditions: consolidated orders, efficient delivery routes, no same-day shipping, and a customer who would otherwise drive alone to a store.
Change any of those conditions and the math changes. Change all of them and online shopping can be more carbon-intensive than a store visit.
The Conditions That Eat the Advantage
Same-day delivery is the clearest example. Standard ground delivery consolidates packages onto optimized routes. Same-day delivery sends a van to your door before the route can be optimized, sometimes with a single package. The carbon cost of speed is rarely communicated to consumers, but it’s real and significant.
Split shipments create the same problem at the order level. An e-commerce company shipping a four-item order in three separate boxes, because items come from different fulfillment centers, has eliminated the consolidation efficiency that makes online retail environmentally competitive. Greenpeace China found that online channels generate 9.4 million tons of packaging material annually; much of that is void-fill in oversized boxes for orders that could have been right-sized.
Returns make this worse. Online return rates run three to five times higher than in-store purchase return rates. Each returned item is re-packaged, re-shipped, and processed through a returns logistics chain that generates its own emissions footprint.
Where the Real Leverage Is
The environmental performance of any e-commerce operation comes down to a small number of operational choices, most of them invisible to customers.
Consolidating fulfillment, shipping all items in a single order from the same location, eliminates split shipment emissions at zero cost to the customer experience. For large retailers with multiple distribution centers, this requires inventory allocation intelligence, but the technology exists and pays back in shipping cost reduction anyway.
Building “slower shipping” as the default (with same-day as an explicit, preferably priced-up option) shifts the fleet efficiency calculation without requiring any customer sacrifice for the majority of orders where timing isn’t urgent.
Ikea discovered that lighter cardboard pallets for online deliveries reduce transportation costs and emissions by 10%. The same logic applies to right-sizing packaging: a shipping carton matched to the product saves weight, volume, and materials simultaneously — reducing emissions and cost in the same motion.
The most sustainable e-commerce company isn’t the one with the greenest marketing. It’s the one that has eliminated split shipments, right-sized its packaging, and made ground shipping the default. None of those changes require the customer to do anything differently.
What Customers Can Actually Do
For consumers who care about the environmental impact of their online purchases, three choices make a real difference:
Order multiple items together rather than placing separate orders as they occur to you. Wait for standard shipping rather than choosing same-day or next-day. Use click-and-collect (picking up an order in-store) when that option is available. It eliminates last-mile residential delivery entirely.
60% of consumers say they would change purchasing habits to reduce environmental impact. The missing piece is usually information: most e-commerce checkouts don’t show consumers the carbon cost of their shipping choice, which means consumers can’t act on the preference they say they have. Retailers who add this information at checkout, even roughly, consistently see higher ground-shipping selection rates.
P.S. If you run an e-commerce operation, the single most impactful change you can likely make for both emissions and operating costs is auditing your split shipment rate. If more than 10% of multi-item orders are shipping in more than one box, that’s both an environmental and financial leak worth fixing.