IoT-enabled pallets and other reusable packaging assets are creating the opportunity to capture real-time data, which can be used to inform and make business decisions more accurately and in a more timely manner than ever before. But feedback loops aren’t new. They have been almost 70 years in development. This article is part of our Stacked Series.
The Produce Department Bullwhip
In my former life at the distribution center, there was a ritual that sometimes played out when the weekly major produce promotion rolled around. Especially if the price was not, and the projected supply of product was not.
At the start of the week, savvy produce department managers at retail stores would inflate their orders for strawberries, cherries, or whatever item was featured in the flyer. The logic was understandable. If supply became tight, they feared their orders would be prorated by upstream buyers. Better to ask for more than they actually need and protect themselves against shortages. And if they did get everything they ordered, it was no problem. Just dial back on the next order.
Meanwhile, the corporate buyers upstream saw the large initial orders and reacted rationally. Demand appeared to be surging, so they secured additional inventory from suppliers. More trucks rolled in. More perishables were staged for delivery.
Then, by Friday or Saturday, as the ad week ended, reality emerged. The inflated demand signal had not reflected true consumer demand at all. Stores were no longer ordering. Buyers found themselves sitting on excess inventory with limited shelf life. Nobody involved was irrational. The structure of the system itself produced the distortion.
This pattern has a name, a history, and an intellectual lineage that many supply chain practitioners have never fully encountered. Understanding that lineage matters now more than ever, because the tools arriving to address it are powerful enough to change the problem entirely, and possibly to create a new one. But we’ll talk about that more in a future installment.
The Man Who Saw the Pattern
Jay Wright Forrester was not a supply chain specialist. He was an MIT engineer who had spent the late 1940s building early digital computers and designing radar defense networks. He understood feedback loops the way most managers understood org charts: intuitively, structurally, and at a level of precision that made ambiguity uncomfortable.
In the late 1950s, Forrester turned his attention to a General Electric appliance factory in Kentucky that was experiencing violent swings in production despite relatively stable consumer demand. Managers assumed the problem was external. Volatile markets, unpredictable customers. Forrester saw something different.
The instability, he argued, was being generated internally, by the structure of the system itself. As information moved upstream through retailers, distributors, warehouses, and factories, delays and distortions accumulated at every step. Every participant added buffer inventory, safety stock, and hedging behavior to protect against uncertainty. By the time demand signals reached the factory floor, they had become magnified echoes of the original consumer purchase.
In 1961, Forrester published Industrial Dynamics, laying out a mathematical framework for how feedback loops, time delays, and information flows combine to produce system-wide instability. His central finding was deeply uncomfortable. Rational, well-intentioned decisions made independently at each level of a supply chain could combine to produce irrational outcomes for the system as a whole. Nobody had to make a mistake for the system to fail.
He called the phenomenon demand amplification. The produce department ritual I described above would have been immediately familiar to him.
Making It Visceral
Forrester’s mathematics were rigorous, but they lived in academic papers. A generation later, MIT’s John Sterman, a student and intellectual heir of Forrester’s, found a way to make the phenomenon visceral.
The Beer Distribution Game, developed at MIT in the 1960s, assigns players to roles in a four-tier supply chain: retailer, wholesaler, distributor, and factory. Each player sees only their own inventory position and incoming orders. Consumer demand is deliberately stable. And yet, without fail, every group of intelligent players produces the same result: wild oscillation, surging inventories, missed orders, and mutual frustration between tiers.
The Beer Game did something that Forrester’s equations could not. It put experienced managers inside the system and let them feel the consequences of their own rational decisions. Sterman’s research into the game’s outcomes added an important dimension to Forrester’s structural analysis. Human psychology was not merely a passive participant in the dynamic. It was an active amplifier. Managers systematically underestimated inventory already in the pipeline, overreacted to apparent shortages, and placed orders that made perfect sense locally while worsening conditions for everyone else.
The structure created the conditions. Human behavior reliably delivered the result.
Naming the Effect
The phenomenon remained largely within academic circles until 1997, when three Stanford researchers gave it the name that stuck. Hau Lee, V. Padmanabhan, and Seungjin Whang published a paper in Harvard Business Review that brought Forrester’s insight to a mainstream management audience. They called it the bullwhip effect, and they identified four specific causes: demand signal processing, the rationing game, order batching, and price variation.
The rationing game, in particular, will sound familiar to anyone who has worked in fresh produce. When supply appears constrained, buyers inflate orders to guarantee adequate allocation, producing the very shortage they feared and feeding a distorted signal upstream. The cause is not dishonesty. It is rational self-protection in a system where the penalty for under-ordering is real, and the penalty for over-ordering is diffuse.
The Lee paper was widely read and recognized. Supply chain executives instantly understood the problem. Academic researchers produced decades of follow-on work. And the underlying problem, the information problem Forrester had identified forty years earlier, remained essentially unsolved.
Decades of Trying
The field did not stand still. From the 1980s onward, a series of technology-driven interventions attempted to address the information delays at the root of the bullwhip effect.
Vendor Managed Inventory transferred replenishment decisions to suppliers, giving them direct visibility into customer inventory levels. In carefully managed partnerships, such as Walmart and Procter and Gamble, being the canonical example, VMI produced real reductions in variability. Collaborative Planning, Forecasting and Replenishment extended the logic further, creating shared forecasting processes between trading partners. Electronic Data Interchange standardized the transmission of order and inventory data. Enterprise Resource Planning systems consolidated internal data flows into unified platforms.
Each intervention was a genuine improvement. Each reduced some portion of the delay and distortion Forrester had identified. And each ran into the same fundamental constraint: they addressed the transmission of information without changing the nature of the information being transmitted.
The signal traveling upstream was still a lagged, batched, order-based signal. It was moving faster, and it was shared more widely, but it was still fundamentally a record of what someone had ordered, not a real-time picture of what consumers were actually doing. The gap between those two things is precisely where the bullwhip lives.
Even in the most celebrated success stories, the limits were acknowledged. A former Walmart executive overseeing CPFR implementation once noted that the initiative never reached its original vision, because trading partners remained too protective of their data to share it fully. The technology was ready. The structural trust problem was not solved.
Why It Persisted
In retrospect, the reason these interventions could not fully resolve the bullwhip effect is straightforward. The information problem Forrester identified in 1958 was not merely difficult to solve with the available tools. It was essentially impossible to solve with them.
Lagged, order-based signals are structurally incapable of eliminating demand amplification. They can reduce it. They cannot remove it. The delay is baked into the medium. Solving the bullwhip effect properly would require something different entirely: not records of orders placed, but real-time observation of demand as it actually occurs. Not reports of inventory positions at the end of a period, but continuous visibility into inventory as it moves. Not forecasts built from order history, but systems capable of detecting shifts in consumer behavior before those shifts have propagated into orders at all.
For most of the period between Forrester and the present, that kind of information infrastructure simply did not exist on a commercial scale. It is beginning to exist now.
A Story That Is Not Finished
Nearly seventy years after Forrester watched a Kentucky appliance factory lurch through cycles it could not explain, supply chain managers around the world are living variations of the same story. The fresh produce ritual I described at the start of this piece plays out, in different forms, across grocery, electronics, pharmaceuticals, and automotive components. The specific products change. The structural dynamic does not.
What is changing, finally, is the information environment itself. Sensors, artificial intelligence, and digital modeling are beginning to provide the kind of real-time, high-fidelity signal that Forrester’s diagnosis implied was necessary. Whether those tools resolve the problem Forrester identified, or simply create a new version of it, is the question the next chapter of supply chain history will answer.
That question is worth a second article of its own. Read our follow-up article on supply chain feedback loops and AI.
This article is based on our white paper: FROM FORRESTER TO FEEDBACK: How AI Is Finally Solving the Bullwhip Problem
Rick LeBlanc, EMBA, is the founder and editor of Reusable Packaging News (subscribe to the free newsletter) and editor of Western Pallet Magazine. A supply chain journalist with more than 30 years of experience, he specializes in pallets, reusable packaging, material handling and related operations. Rick is co-author of Pallets: A North American Perspective and Pallets & Progress: A Collected History of Pallets and Palletized Handling 1922–1945. He is also an advisory board member of the Virginia Tech Center for Packaging and Unit Load Design. Read Rick’s full bio.
