Fancy, No. 1, No. 2, Cull: The Four-Tier System a 1912 Produce Grader Used to Route Their Produce

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.

On a cool morning in the fall of 1912, at the Washington Street produce terminal in lower Manhattan, a grader would work his way down a row of crates stacked twelve high along the dock. He moved quickly, pulling lids, pressing thumbs into flesh, lifting fruit to the light coming through the clerestory windows above. Each crate got a mark: Fancy, No. 1, No. 2, or Cull. The mark was written in chalk on the side of the crate, and in a ledger that traveled with the shipment back to the commission merchant who had brokered the load.

The grader was not a quality inspector in the modern sense. He was a router. The mark he applied did not only describe the fruit, it determined where the fruit was going and what it would earn. Fancy apples, the kind with uniform color, no blemishes, and firm flesh all the way through, were destined for the display cases of fine grocers and hotels. They earned the highest price per barrel and were handled with corresponding care. No. 1 apples, sound but slightly irregular in size or color, went to general retailers across the city. No. 2 apples, still edible but bruised or undersized, moved to smaller neighborhood vendors or to early processing operations, where they would become applesauce or cider. Culls — overripe, cracked, damaged — went to livestock feed, or in some seasons were sold in bulk to distillers or vinegar producers for almost nothing. Nothing left the dock without a destination, and every destination had a margin attached to it.

This system predated the USDA's first formal grading standards by several years. The agency had established its Office of Markets in 1913 and would issue the first official federal grade standards for potatoes in 1917, driven by wartime pressure to standardize agricultural supply chains at scale. But the terminal markets had been running tiered grading systems through private convention long before federal standards arrived. Commission merchants, wholesale receivers, and growers had all agreed — informally but consistently — that a product's grade was not a judgment call. It was an operational fact that triggered a specific set of downstream decisions: which buyer, which price, which truck, which channel.

The terminal markets themselves were built for this logic. At the Wallabout Market in Brooklyn, which by the early twentieth century had become one of the largest wholesale food distribution centers in the world, the physical layout separated incoming product by grade in ways that made the routing nearly automatic. Buyers for hotels and institutions moved through different sections than buyers for street vendors. Truckers for canneries loaded at separate docks. The grading decision made at the inspection point cascaded through every subsequent step without anyone having to reconsider it.

A grader's speed mattered as much as his accuracy. Produce depreciates by the hour. An apple that is No. 2 today may be Cull by tomorrow morning, which means a lower price, a different destination, and a narrower margin — sometimes no margin at all. The people who built these systems understood that grading throughput and value recovery were not separate concerns. They were the same concern.

What the produce terminal operators had built, well before anyone called it supply chain management, was a disposition framework with a routing engine embedded in it. The grade was the data point. The destination was the output. The speed of classification was the variable that controlled how much value survived the handling process.

This is worth slowing down to examine, because it runs counter to how most modern operations think about condition assessment. In a typical returns workflow today, grading and routing are treated as sequential decisions: first we figure out what grade the item is, then we figure out where it should go. The produce system collapsed those two questions into a single act. The grader did not assess quality and then consult a routing table. The grade was the routing table. The terminology itself encoded the destination: Fancy meant premium retail, Cull meant the lowest-value channel available.

For each grade tier, the system had a clearly defined cost-to-serve associated with it. Fancy product required careful stacking, protective wrapping, and delivery to specific accounts with time windows. No. 2 product moved in bulk with less handling and reached buyers who were optimizing for price, not condition. Culls went where they could go quickly, because holding them was more expensive than moving them at low margin. The entire structure was calibrated around the insight that holding cost and processing cost are inversely related to grade — and that the longer a lower-grade item sits, the worse that relationship gets.

The USDA's formalization of these standards in 1917 and beyond was not the creation of the grading idea. It was the standardization of a logic that operators had already proven worked at scale. The federal standards allowed that logic to travel — across states, across commodity types, across buyers and sellers who had no prior relationship. A crate marked U.S. No. 1 in California meant the same thing to a receiver in New York because both parties had agreed to the same definition of what that grade required and what it implied about destination and price. The grade was, in effect, a common language for disposition decisions made at high speed across a distributed network.

The produce grading system evolved. Federal standards expanded through the middle of the twentieth century to cover hundreds of commodity types. The Agricultural Marketing Act of 1946 gave the USDA formal authority to standardize quality grades and offer voluntary inspection services, institutionalizing what terminal markets had been doing through convention for decades. Today, the AMS administers grade standards for everything from fresh apples to canned tomatoes, and the grading logic remains intact: a defined tier maps to a defined channel with a defined margin expectation.

What did not make the journey into modern operations was the discipline that made the original produce system work. Ecommerce returns, which accounted for roughly $890 billion in returned goods in 2024 alone, are processed in environments where the connection between grade and route is rarely as clean as it was on those terminal market docks. The average cost to process a return runs between 20 and 65 percent of the item's original value, depending on category and handling requirements. That spread is enormous, and much of it comes from a single failure: the disposition decision gets made too late, too slowly, or too inconsistently to preserve the value that faster classification would have captured.

A returned item that sits in a receiving area is the equivalent of a crate of apples sitting on the dock overnight. The fruit does not become more valuable by waiting. And yet most returns operations — by design or by accident — treat the inspection and disposition steps as separate workflows rather than a single integrated act. Items are received, logged, queued, inspected later, routed later still, and by the time they reach their final channel, a meaningful fraction of their recoverable value has elapsed.

The operators who built the terminal market system did not have the luxury of this delay. They were moving perishables through a network with no tolerance for dwell time, and they had built their classification system around that constraint. The grade went on the crate at the point of first inspection, not after a series of handoffs. The routing decision was made once, at the earliest possible moment, by someone whose job was to make it fast and make it right.

Modern reverse logistics has the data, the systems, and the labor capacity to do this. What it often lacks is the conviction that the grade and the routing decision are the same decision, made at the same moment, by the same person with the authority to act on it. The produce grader did not write a recommendation. He wrote a destination.

There is also the matter of what happened to the culls. The terminal market system was not sentimental about low-grade product — it moved culls out fast, at whatever price the channel would bear, because zero margin is better than negative margin and storage cost is always accumulating. The willingness to route aggressively to lower-value channels, without trying to regrade or rework every item into a higher tier, was a discipline that kept the overall system solvent. Modern returns operations frequently invert this logic, spending more on rework and refurbishment than the incremental recovery value justifies, while faster and cheaper disposition paths sit underutilized.

The produce graders understood that not every item can be Fancy, and that trying to make a No. 2 into a No. 1 costs more than the price difference. They also understood that a Cull sitting on the dock is not a recovery opportunity — it is a cost that compounds by the hour.

For practioners: Where in your current returns operation is the disposition decision being made, and how far downstream from the first point of inspection does it actually happen? If your grading step and your routing step are separated by time, handoffs, or different teams, what is the value that disappears in that gap? Do you have an RMS that clearly defines grade tiers with predetermined destination channels — not recommendations, but rules — so that the person doing the inspection does not have to make any decisions? And for the lowest-grade items in your network, are you moving them as fast as the culls moved off the terminal market docks, or are they accumulating dwell time while you look for a better outcome that the grade does not support?

Thank you for reading Sunday Supply Chain Stories!

If you enjoyed the newsletter, follow us on LinkedIn and ask your friends to subscribe to receive the next story on Sunday!