WW/PROCUREMEN
A 20% Finished-Goods Buffer Beats Raw-Material Stock by 1.4x
An MIT-modeled mattress supply chain study finds finished-goods buffers 1.4x more effective than raw material at protecting revenue during sole-source failures.
- By
- Elena Vasquez
- Filed
- Length
- 711 words
- Read
- 4 min

Key points05
- A 20% finished-goods safety stock increase cut lost sales by about 75%, versus 53% for the same raw-material increase.
- Finished-goods buffers proved about 1.4 times more effective than raw material at protecting revenue.
- In a nine-week outage, two extra weeks of finished goods avoided nearly three times the inventory's cost in lost sales; service levels rose about 4%.
- A multi-sourced raw material disruption caused no lost sales; a sole-source cutoff drove rapid service decline once buffers dried up.
- The MIT CTL study used a Mixed-Integer Linear Programming model simulating a full year of U.S. supply chain operations.
A 20% increase in finished-goods safety stock cuts lost sales by roughly 75% during a sole-source supply failure, while the same increase held as raw material delivers only a 53% reduction. That is the headline finding of an MIT Center for Transportation and Logistics study that modeled the U.S. supply chain of a leading premium mattress maker, and it gives procurement and planning teams a rare like-for-like comparison of where a resilience dollar works hardest.
The thesis, Optimal Risk-Mitigation Investments in Tempur-Pedic's Mattress Supply Chain Network, was authored by Rishi Raghav Sanker and Jiachen (Jessica) Mo under the supervision of Dr. Chris Mejia Argueta, research scientist at MIT's Emerging Market Economies Logistics Lab. The researchers built a Mixed-Integer Linear Programming model tracing flows from suppliers through plants to distribution centers, then stress-tested the network against three failure modes over a full simulated year of operations.
The company's flagship line, with premium models selling for thousands of dollars per unit, drives substantial margins. Resilience spending, however, had grown fragmented across inventory, production capacity and logistics flexibility — all competing for one limited budget. Buffering everything locks up working capital; buffering in the wrong places leaves real risks exposed.
How did the network handle each disruption?
The first scenario shut a plant for three, six and nine weeks. The network proved surprisingly robust: other plants absorbed the lost output and finished-goods inventory covered the gap, so customers saw no lost sales. The resilience came at a price, though. Inventory drew down sharply and inter-facility transfers spiked — the system coped through execution effort rather than built-in redundancy.
The second scenario disrupted a critical raw material sourced from several vendors. The network barely registered the shock. Backup suppliers stepped in, avoiding lost sales or meaningful added cost, confirming that existing multi-sourcing already provided adequate coverage.
The third scenario exposed the network's true weak point. When the model cut off a material with only one supplier, finished-goods buffers ran dry, service levels slid, and lost sales climbed quickly. This is where the buffer trade-offs became measurable.
What does the buffer comparison show?
Two defensive options were compared: holding more raw material upstream versus holding more finished goods downstream. Both helped — but not equally. Raw-material buffers reduced lost sales, yet their benefit faded fast; beyond a 20% increase, extra material added little value while pushing total costs higher.
Finished goods told a different story. In the nine-week outage case, two additional weeks of finished-goods inventory avoided nearly three times the inventory's cost in lost sales. Service levels also rose about 4%, and the finished-goods increase even trimmed total network cost slightly by softening downstream impact. Across the analysis, finished goods proved about 1.4 times more effective at protecting revenue than an equivalent raw-material increase.
What should shippers and planners take from it?
The study recommends a differentiated strategy built on five priorities, with finished goods as the broad first line of defense — in practice, holding them at about 1.2 times current levels while keeping raw-material buffers lean. Selective raw-material protection should target single-sourced, business-critical inputs, where the model showed the sharpest revenue risk.
Inventory alone, the authors stress, is not enough. Execution capabilities — flexible outbound logistics, dynamic reallocation across distribution centers, and flexible sourcing across plants — determine whether buffer stock actually converts into preserved customer service. In the plant-shutdown scenario, it was precisely those reallocation and transfer mechanisms that prevented lost sales.
The commercial logic extends beyond mattresses. Any manufacturer weighing where to place safety stock against finite working capital can replicate the approach: map the network, simulate realistic failure modes, and compare mitigation investments side by side within real constraints such as plant capacity, storage space and inventory rules. For networks carrying premium, high-margin products, the cost asymmetry between a modest finished-goods buffer and a stockout makes the case on its own.
The study's broader trajectory points toward precision over volume in resilience spending: as disruption modeling becomes standard practice, expect more shippers to shift budgets from blanket upstream stock toward targeted downstream buffers and the flexible logistics needed to activate them.
Source: scmr.com
More from Elena Vasquez
Show full bio
News editor covering people, launches and day-to-day practice in Ocean and air freight, container shipping at Waybill Wire.
929 articles