For retailers, counting stock is a chore and requires a sizable investment in labor. Aside from annual audit requirements, is stock counting really necessary? Are accurate inventory records really that important? Since 2018 my role at Datascan has had me exploring the concept of inventory record accuracy. I’ve long searched for a reputable study on the sales impact of counting more often in a non-RFID environment, and it finally came out in late 2019. The ECR Community Shrink & OSA Group funded an academic study titled “Inventory Inaccuracy in Retailing: Does it Matter?” by Prof. Dr. Yacine Rekik (emlyon business school, France), Prof. Dr. Aris A. Syntetos (Cardiff Business School, Cardiff University, United Kingdom), and Prof. Dr. Christoph H. Glock (TU Darmstadt, Germany). I highly recommend you read this study for more details - it is certainly the largest study on this subject. Read on for the highlights.
High-level summary:
- The study covered about 1 million SKUs in 100 stores among 7 grocery and apparel retailers in Europe
- 60% of the SKUs were found to be inaccurate
- The average discrepancy was +6.6 units for overages and -6.0 units for shortages
- Correcting the inaccurate stock levels leads to approximately 4% to 8% increased sales
- The biggest opportunity comes from higher-volume, more expensive items
The authors note there are only 3 broad actions a retailer may take to address inaccurate inventory records:
- Implement preventive measures to avoid inaccurate inventory records
- Implement corrective measures to fix inaccurate inventory records
- Use an inventory management system to comprehend the expected inaccuracies and replenish accordingly
The study focuses on #2, the impact of correcting inaccurate records by counting more frequently. Counting is the only way to correct inaccurate inventory records, regardless of the root cause. Retailers should continue to implement preventive measures, but these will never completely eliminate inaccuracies. The authors note that #3 (incorporate inaccuracies into replenishment algorithms) does not appear to be available at this time. The authors performed two experiments: one measures the impact on sales for all seven retailers and the other measures the rate of decline in inventory accuracy for one retailer.
A total of seven retailers in grocery and apparel participated in the study: four in grocery/general and in three fashion/apparel. Stores were divided into test and control groups after being matched by store location type, store size, management, and technological capabilities. Of the approximately one million SKUs between the retailers, the study used about 233,000 for reaching conclusions. To be included the SKUs had to be sold both before and after the additional stock audit in the test stores, counted in the stock audit(s), and belong to the same class of fast, medium or slow movers between the test and control stores. The study performed an initial and final stock count for all stores, and an intermediate stock count for the test stores. The control stores did not perform an intermediate stock count. The experiment lasted 24 weeks, and sales for the last 12 week period were compared between the test and control groups.
The additional count for the test stores resulted in approximately 4% to 8% in increased sales for the last 12 week period. In fact, looking at only the fast movers with high discrepancies, they found a 14% increase in sales for the same period. I will explore how more accurate inventory records may contribute to increased sales in future articles. This chart demonstrates when the retailer should expect increased sales as a result of cutting the time between counts in half. Increased sales occur between counts 2 and 3 and between counts 4 and 5. The red lines show how far inventory record accuracy degrades from a less frequent count schedule. The green areas represent the improved inventory record accuracy as a result of the additional counts. There is a clear sales lift resulting from additional counts that correct inventory record inaccuracies for both overages and shortages. The authors strongly suggest that retail leaders consider the frequency and timing of stock counts to be a critical strategic initiative to increase sales, especially for fast movers with high discrepancies.
Want to repeat this experiment in your own stores? The study’s appendix contains the steps necessary to replicate the experiment.
