This is Sunday Supply Chain Stories, where we revisit the foundations that continue to shape how inventory moves, returns and recovers value.
If you enjoy the newsletter, follow The Supply Loop on LinkedIn and ask your friends to subscribe to receive the next story.
Every morning in the early 1950s, the milkman arrived at your doorstep before you were awake. They delivered full glass bottles and collected the empties left out from the previous delivery.
In that single moment, two logistics problems were solved on one truck.
At that time, 94% of all milk consumed in the United Kingdom and United States moved through this system. Each glass bottle made the trip 25 to 30 times on average. The bottle was not treated as disposable packaging. It was an asset.
The route design was the operating principle: one delivery path meant empty bottles moved back on the same truck, along the same roads. At the dairy, the process was straightforward. Inspect, wash, sterilize, refill, circulate. The container moved with the product continuously, without introducing new packaging at each cycle.
Collection happened as part of the process, not as a separate event. The customer's role was minimal. Leave the empty on the step. The structure absorbed the rest.
When forward and reverse flows share the same route, recovery becomes part of the delivery. The cost to bring the asset back is absorbed into the network that already exists. That changed the unit economics entirely.
Why did this work? Because a returnable bottle operates across multiple cycles. When a bottle circulates 25 to 30 times, the manufacturing cost is distributed across those uses. A 2022 lifecycle analysis found that reusable glass bottles become environmentally preferable to single-use HDPE plastic after just 1.7 uses. Compared to beverage cartons, they break even after 2.0 uses.
The value was not captured in one sale. It was realized through repeated circulation.
By the 1970s, the model had already begun to collapse.
Why did it stop? Home refrigeration meant people could shop less frequently. Increased car ownership made the supermarket trip convenient. Delivered milk declined from 30% of consumption in 1963 to under 7% by 1976, and below 1% by the 1990s.
The economics flipped. Single-use cartons and plastic jugs became the dominant format. They required no recovery infrastructure, but they produced significant waste.
The packaging shifted from asset to expense. Once it was treated as an expense, it moved in one direction.
This structure should be familiar to anyone working in returns or ecommerce. Most reverse logistics challenges come from how the system is designed: forward and reverse flows operate separately, supported by different infrastructure, different cost structures, and different decision points.
The milkman operated in a different model. Delivery and recovery were the same operation.
Today, carriers arrive at homes and businesses every day at predictable frequency and reach. Parcel carriers, grocery delivery services, meal kit companies, rental services all manage outbound networks of significant sophistication.
But what moves back through that same system? In most cases, very little. Returns are managed separately, with different costs, different logistics, different decision points. A package arrives. It sits. It eventually moves toward a returns center through a different network, often weeks later.
Empty bottles, reusable containers, packaging materials could move back through the same network that delivered them. Not as a secondary process, but as an integral part of the route design.
The manufacturing cost is the largest environmental burden in any packaging system. For reusable glass, cleaning and distribution add secondary impacts, but when a bottle is used 25 times instead of once, that production burden is amortized across 25 journeys. The return logistics becomes the smaller variable.
The carrier that already visits your location has the opportunity to solve two problems at once: deliver what you ordered and recover what you no longer need.
For practitioners: If you modeled your highest-return SKUs across a second and third lifecycle, how would the margin profile change? Where in your network could forward and reverse flows realistically share the same path today? What percentage of your returns are treated as assets versus costs?
Sources
United States Department of Agriculture. Fluid Milk Delivery Statistics, 1963 and 1976.
Gamut Packaging. "When Did the Milkman Stop Delivering Milk." February 2025.
The Modern Milkman / Friends of Glass UK. Life Cycle Analysis of Glass Milk Bottles. 2022.
Thank you for reading Sunday Supply Chain Stories!