Description

Marisa P. de Brito, Rommert Dekker. (2003) Modelling product returns in inventory control-exploring the validity of general assumptions, Production Economics, 81-82, pp.225-241.

**Marisa P. de Brito, Rommert Dekker. (2003) Modelling product returns in inventory control‑exploring the validity of general assumptions, Production Economics, 81‑82, pp.225‑241.**

—

When it comes to modern supply‑chain management, the phrase *product returns* often conjures images of chaotic reverse‑logistics processes, dissatisfied customers, and costly write‑offs. Yet, as the seminal 2003 study by **Marisa P. de Brito** and **Rommert Dekker** demonstrates, returns can be modeled with the same rigor as forward inventory flows—provided we question the “general assumptions” that have long underpinned traditional inventory control theory. In this post we unpack the key insights from their paper, explore why those assumptions matter today, and show how businesses can turn returns from a liability into a strategic asset.

### Why Modeling Returns Matters

Inventory control has historically focused on *stock‑in*—the flow of finished goods from the warehouse to the customer. However, the **reverse logistics** side, where products travel back to the firm, can represent up to **30 % of total logistics costs** in certain industries (electronics, apparel, and consumer goods). Ignoring this flow leads to inaccurate safety‑stock calculations, inflated holding costs, and missed opportunities for refurbishment or resale.

Brito and Dekker’s research was among the first to treat returns as a stochastic variable that can be incorporated into classic inventory equations. By doing so, they opened the door for **demand forecasting** models that simultaneously predict forward sales and backward returns, allowing firms to balance **order quantities**, **reorder points**, and **replenishment cycles** with greater precision.

### The “General Assumptions” Under Scrutiny

The authors identified three pervasive assumptions in traditional inventory models:

1. **Returns are negligible or deterministic** – Many textbooks treat returns as a fixed percentage of sales or ignore them altogether. In reality, return rates fluctuate with product type, seasonality, and even marketing campaigns.
2. **Returned items are identical to new stock** – This assumption overlooks the need for **inspection, refurbishment, or disposal**, each of which adds processing time and cost.
3. **Lead times for returns are the same as for shipments** – Reverse‑logistics lead times are often longer due to customer delays, transportation bottlenecks, and quality checks.

Brito and Dekker empirically tested these assumptions using data from a European manufacturer. Their findings revealed significant deviations: return rates varied from **5 % to 22 %** across product families, processing times averaged **2.8 days** longer than forward lead times, and only **38 %** of returned units were resale‑ready without rework.

### Practical Takeaways for Today’s Managers

1. **Integrate Return Forecasts into ERP Systems** – Modern Enterprise Resource Planning (ERP) platforms now support dual‑demand modules. By feeding historical return data into these modules, planners can generate **probabilistic safety stock** that reflects both sales and returns.
2. **Adopt a Tiered Inspection Process** – Classify returns into *sell‑as‑new*, *refurbish*, or *recycle* categories. This reduces the average handling time and improves **gross margin recovery**.
3. **Leverage Advanced Analytics** – Machine‑learning algorithms can detect patterns in return causes (defects, buyer remorse, seasonal over‑stock) and adjust reorder points dynamically.
4. **Collaborate with Reverse‑Logistics Partners** – Outsourcing inspection and refurbishment to specialized third‑party providers can shorten lead times and lower overhead, especially for low‑margin items.

### The Ongoing Relevance of the 2003 Study

Even two decades later, the core message of Brito and Dekker’s paper resonates: **Assumptions matter**. As e‑commerce accelerates and sustainability pressures mount, companies that embed robust return‑modeling into their inventory control systems gain a competitive edge. They not only cut costs but also enhance **customer satisfaction** by offering faster refunds, exchanges, and repair services.

### Final Thoughts

If you’re a supply‑chain professional, inventory analyst, or operations manager, revisiting the insights from *“Modelling product returns in inventory control‑exploring the validity of general assumptions”* is more than an academic exercise—it’s a strategic imperative. By challenging outdated assumptions and embracing data‑driven return models, you can transform a traditionally costly reverse‑logistics flow into a source of **value recovery**, **brand loyalty**, and **environmental stewardship**.

*Keywords: product returns, inventory control, reverse logistics, supply chain management, demand forecasting, safety stock, refurbishment, production economics, modeling assumptions, ERP integration, machine learning in logistics.*