Supermarket Pull Systems Part II: Turning Principles into Practices

Editorial Note

This article is Part 2 of a three-part series on supermarket pull systems.

Part 1 focused on why supermarket pull works: fixed locations, clear limits, and replenishment triggered by actual consumption. It showed how supermarkets create stability by placing and governing inventory rather than relying on forecasts or schedules.

Part 2 examines how those same rules hold up in practice. Using recent case studies across manufacturing, healthcare, retail, and distribution, it looks at what happens when supermarket limits are respected—and what happens when inventory grows beyond them.

Review

In Part 1, we described supermarket pull as a system governed by simple rules that tie replenishment directly to use. When something is taken, it is replaced. When limits are reached, the system signals a problem instead of hiding it.

That logic is easy to understand. Sustaining it once the system is running is far more difficult.

Recent case studies show a consistent pattern.

In manufacturing, workers initially produced “just in case,” creating duplicate inventory alongside new kanban systems during transition (Martins et al., 2021).

In healthcare, supply closets accumulated inventory beyond kanban limits when leadership reinforcement faded during the first 90 days (BMJ Open Quality, 2024).

In retail, warehouse layouts designed for push replenishment created problems that persisted even after pull procedures were introduced (Vieira et al., 2025).

Supermarket pull systems rarely fail technically. The kanban signals remain. The WIP limits stay posted. What erodes is the discipline to follow the rules when the inevitable pressure of operational realities rears its’ ugly head.

Supermarket Pull Works—But Only Under Specific Conditions

Before we get into the evidence, it’s worth being precise about where supermarket pull fits and where a different pull method is the better choice.

A supermarket works only under a clear condition: replenishment must reliably respond faster than the downstream process consumes what is on the shelf.

When that condition holds, the discipline of the supermarket takes effect: fixed locations, clear limits, and replenishment triggered by actual use create stability and surface problems at a manageable pace.

When that condition does not hold, the supermarket does not improve the system. It redistributes the pressure. Stockouts become routine. Expediting becomes normal. Inventory limits expand, creating the illusion of control while obscuring the real constraint.

In those situations, pull still matters, but the pull design must reflect the true response time of the work.

Research on supply chain strategies show predictable breakdowns

• When lead time exceeds consumption frequency. Pull signals arrive too late to be useful. Strategic push inventory outperforms reactive replenishment.

• When customer demand is sporadic. Low-frequency or irregular consumption does not generate reliable pull signals.

• When setup economics dominate. High setup/changeover time and associated costs force batch sizes beyond reasonable supermarket limits unless SMED is addressed first.

• When upstream variability must be buffered. Hybrid strategies (push upstream, pull downstream) often outperform pure pull in complex networks.

Key Takeaway

Supermarket pull is not an inventory optimization technique. It is a distinct operating system, governed by specific conditions and prone to predictable failure modes when those conditions are ignored.

What Supermarket Discipline Actually Means

A shared office kitchen has a coffee shelf holding exactly three containers of beans. When the last container is taken, the empty spot signals: reorder one.

If coffee runs out before replacement arrives, the problem is visible. Consumption increased, replenishment lagged, or the limit was wrong. The response is not to double-order next time, but to understand what changed and adjust the system.

This exposes the central tension:

When should supermarket limits be adjusted versus defended?

Legitimate adjustments respond to sustained change:

Lead time increases, consumption doubles and holds, yield drops.

What undermines supermarket pull is reactive expansion:

Adding inventory in response to a stockout without understanding whether the limit was wrong or the system failed to perform.

That forcing distinction is the value (and difficulty) of supermarket systems.

Manufacturing: When Capability Must Match Requirements

A 2021 case study of a European automotive components manufacturer documented the shift from forecast-driven push to supermarket-governed pull (Martins et al., 2021).

Results included:

• 56% reduction in WIP between injection and painting.

• 45% reduction in WIP between painting and expedition.

• 38% reduction in setup times.

• Significant lead-time reduction.

The gains came from removing discretionary production and letting supermarket limits regulate flow.

What challenged the system was capability.

Operators initially produced ahead because setup times made small batches economically painful. Once SMED reduced setup time by 38%, the economics shifted. Producing to replace consumption became viable, and excess WIP disappeared.

Key Takeaway

Supermarket discipline requires system capability to respond. When capability lags, limits will be violated—not because people lack discipline, but because following the rules produces worse outcomes than breaking them.

Healthcare: When Leadership Makes Compliance Visible

A 2024 BMJ Open Quality study reported that converting a central hospital storeroom to a supermarket Kanban system achieved:

• 47% reduction in weekly supply costs.

• Near elimination of stockouts.

• Reduced nursing time spent ordering.

• Improved staff satisfaction

A separate pilot using automated supermarket Kanban eliminated manual nurse ordering and reduced total order volume by more than 50% (Identi Medical Systems, 2025).

What challenged the system was reinforcement. During busy periods, staff reverted to “just in case” ordering, accumulating inventory beyond limits.

The turnaround came from daily gemba walks by nurse managers auditing kanban compliance. Two questions were asked:

• Does inventory match the kanban limit?

• If not, what prevented the system from working?

Violations were treated as system data, not personal failure. Within 90 days, noncompliance dropped from 30% of positions to under 5%.

Retail: When Infrastructure Fights the System

A 2025 case study of a European grocery retailer examined high-volume stores shifting to supermarket replenishment triggered by shelf consumption (Vieira et al., 2025). Results included:

• 32% reduction in daily labor hours.

• 64% improvement in product availability.

• 15% reduction in replenishment wait time.

What challenged the system was infrastructure. Warehouses designed for push replenishment relied on large pallets and batch breakdowns.

Under supermarket pull, products needed immediately were trapped upstream. Ergonomic trolleys improved flow but could not fully overcome layouts designed for push.

Key Takeaway

Supermarket discipline depends not just on people and processes, but on whether physical infrastructure supports the operating system.

Distribution: When Limits Must Match Measured Reality

A 2024 case study of a fresh and frozen goods distribution center applied supermarkets with protected FIFO lanes and replenishment triggered by downstream demand (Chiaraviglio et al., 2024). Results included shorter lead times, reduced congestion, and lower spoilage.

What challenged the system was assumed lead time. Buffers sized “to be safe” failed when variation exceeded expectations. Larger buffers then worsened FIFO access, trapping older inventory behind newer arrivals.

When the operation sized limits using measured lead-time distributions rather than assumptions, reliability improved and spoilage declined.

Three Mechanisms That Build Supermarket Discipline

Across cases, supermarket discipline usually holds when three mechanisms are aligned:

1. Capability matched requirements. Pull signals were economically feasible to follow.

2. Leadership made compliance visible and daily. Violations were treated as diagnostic data.

3. Infrastructure supported the operating system. Layouts and handling matched pull flow.

Why This Matters Now

AI and machine learning increasingly optimize demand forecasts and inventory parameters. As Harvard Business Review notes, these systems reinforce existing operating behavior (Davenport & Ronanki, 2023).

Without supermarket discipline, AI optimizes around dysfunction

Better safety-stock math, smarter expediting, more accurate forecasts of instability caused by batching.

With governed limits and replenishment tied to use, AI enhances performance within boundaries: detecting lead-time shifts, identifying instability, and supporting adjustment rather than masking problems.

AI optimizes the system you give it. Supermarket discipline ensures you are optimizing the right system.

Summary

Part 1 established why supermarket pull works. Part 2 examined how it performs when tested in practice.

The evidence is consistent. Supermarket pull succeeds when capability, leadership, and infrastructure support its rules. It degrades when inventory exceeds its limits—not because people lack discipline, but because the system cannot sustain it.

The logic is simple.

Maintaining the conditions that make it viable is the work.

Citations & Further Reading

Manufacturing

Martins, B., Silva, C., Silva, D., Machado, L., Brás, M., Oliveira, R., Carvalho, T., Silva, V., & Lima, R. M. (2021). Implementation of a pull system: A case study of a polymeric production system for the automotive industry. Management Systems in Production Engineering, 29(4), 253–259.

https://doi.org/10.2478/mspe-2021-0031

Healthcare

BMJ Open Quality. (2024). Lean Kanban implementation in hospital supply management. BMJ Open Quality, 13(1), e002388.

https://bmjopenquality.bmj.com/content/13/1/e002388

Identi Medical Systems. (2025). Automated Kanban and PAR replenishment in hospital wards: Case studies.

https://identimedical.com/resources/case-studies/

Retail & Distribution

Vieira, E., Tomaz, L., Leitão, J., Fernandes, J., & Dinis-Carvalho, J. (2025). Enhancement of in-store product replenishment flow and introduction of pull approach in a food retail chain. Logistics, 9(2), 61.

https://doi.org/10.3390/logistics9020061

Chiaraviglio, A., Grimaldi, S., Zenezini, G., & Rafele, C. (2024). Overall warehouse effectiveness (OWE): An integrated performance indicator for warehouse operations. Logistics, 9(1), 7.

https://doi.org/10.3390/logistics9010007

Inventory Strategy & Pull System Theory

Boonmee, C., Arimura, M., & Asada, T. (2020). Push versus pull inventory strategies under demand uncertainty. European Journal of Operational Research, 285(3), 854–867.

https://doi.org/10.1016/j.ejor.2020.03.042

AI, Analytics, and Operations

Davenport, T. H., & Ronanki, R. (2023). AI in supply chain management. Harvard Business Review.

https://hbr.org/2023/07/ai-in-supply-chain-management