The use of technology in retail is not new. However, in recent years we have seen an increase in its benefits and possibilities within stores. The digital world is here to stay, and retailers must be at the forefront to make the most of these new tools that have become indispensable in decision making, process improvement, and taking advantage of the opportunities it offers. 

The problem of supply is a reality for all retailers, which goes beyond customer dissatisfaction and lost sales. Companies make great efforts to design products, strategies, and campaigns. However, this effort will be in vain if the product is not available on the shelf for the customer. 

In this context, retailers are paying more and more attention to inventory management and the replenishment process. But what about managing phantom inventory?

Phantom inventory is when the system shows positive stock for an SKU or item, but the store, in reality, has no units of that SKU, or the product is not visible for the customer to purchase. In other words, it refers to products that the system thinks are available to the customer, but in practice, they are not. They are like ghosts since they appear in the inventory system but do not exist on the shelf. 

These discrepancies between the actual inventory and the one reported in the system can arise for several reasons. According to a study conducted by MIT Supply Chain Management, among the most common causes are the following:  

1. Inventory errors in the sale: mistakes can be generated when scanning the products when making the sale. This is a common situation in products with varieties, such as different yogurt flavors. 

2. Errors in receiving inventory: are generated when products are received without a detailed review. For example, the entry of products in pallet format or grouped instead of considering the unit. In these cases, the discrepancy can be generated when entering and confirming a quantity different from the physical amount in the system.  

3. Operational execution and compliance: Non-compliance with procedures and protocols can lead to phantom inventory problems when, for example, products are left in the warehouse or backroom and not on the shelf. Products are left out of sight in the warehouse. It can also happen due to damaged or shortage products that were not recorded in the system.

4. Shrinkage: inventory reduction can be caused by loss, theft, damage, or product consumption inside the store or warehouse by employees or customers. 

5. Misplaced products: both store personnel and customers can cause some products to be misplaced. It can happen because products are stored or displayed on another shelf, hidden behind similar products, or moved from their original place by customers inside the store. 

The impact of phantom inventory in retailing 

In our previous blog, we discussed the impact of out-of-stocks on the final sale, introducing the term Out-Of-Stock (OOS) and On-Shelf Availability (OSA). 

Retailers have visibility of stock through their inventory management systems. OOS detection is done from the same system, as it contains information on SKUs that have stock equal to or less than zero. 

Studies show that, in general, the inventory management system reveals stock problems for 2% of SKUs. This stock-out level could be considered the tip of the iceberg, as studies by the Grocery Manufacturing Association (GMA) and Efficient Consumer Response Europe (ECR Europe) show that the stock-out level is 8% in both the United States and Europe. 

The challenge of phantom inventory lies here, as 75% of on-shelf stock-outs are not reflected in the system. 47% of cases are due to a lack of stock at the store or backroom (hence, the inventory level recorded in the system does not match the stores’ reality). In the remaining 53% of cases, the problem arises because the store has inventory, but the product is not available within reach of the customer for sale.

According to the same GMA study, a retailer faces a 43% loss of potential direct sales when no stock is available on the shelf. Thus, the impact of stock-outs due to phantom inventory in retail would amount to a direct loss of 2.6% of sales. 

AI to identify and manage phantom inventory

The use of artificial intelligence (AI) in retail is on the rise. According to studies by Juniper Research, investment in this sector will reach $7.3 trillion by 2022 in areas such as customer service, automated marketing, and demand forecasting. This projection could be exceeded given retail’s ability to adapt to pandemic-driven changes in consumer habits. 

According to Boston Consulting Group (BCG), the increased use of AI in retail is critical to driving business revenue. AI allows retailers to generate and evaluate forecasts considering variables in several scenarios, such as demand, supply, inventory, pricing, and logistics. In addition, thanks to this technology, large databases can be analyzed in a short time, boosting decision-making based on real-time information. 

The development of technology has enabled the incorporation of AI-based replenishment systems. As we saw earlier, the causes of phantom inventory are varied, making its management a complex task. Given this, AI has focused on analyzing the symptoms of phantom inventory, such as sales anomalies. 

AI systems have significant advantages, as they perform a statistical analysis to evaluate the probability of an SKU being out of stock. These systems integrate a high number of variables and Machine Learning algorithms, which identify data patterns. 

Based on this analysis, it is possible to detect if the sale of a product is anomalous. We speak of anomalous sales when the actual sales of a specific SKU do not match what was planned by the demand projection system. Under AI analysis, it is unlikely that these products have zero sales if they are available on the shelf, so it is inferred that there are phantom inventory problems. 

Furthermore, a notification system can be added to the probabilistic analysis. Then, an automatic alarm will be triggered to tell store personnel which SKUs have phantom inventory problems and should be checked in the backroom for shelf replenishment.

Frogmi’s global vision goes beyond problem reporting and alerts. Frogmi offers a complete SKU-level product management solution that considers manual replenishment, specific notifications, and SKU-level tasks generated by artificial intelligence. Thanks to Frogmi’s integrated solution, the information does not remain stagnant in a report or email but is taken directly to the stores to be managed. The ultimate goal is to have a complete vision that allows preventing, identifying, and addressing possible stock-outs as soon as possible to deliver the best availability and increase sales.

At Frogmi, we know that the implementation of task management tools integrated with technology and AI solutions are welcomed by store personnel, as they deliver valuable information oriented to execution with concrete objectives. 

Although it often goes unnoticed, phantom inventory comes at a high cost to retailers. Fortunately, the technology to identify and manage it is already available through AI and in-store operation platforms. All that remains is to join this trend, so phantom inventory problems become a thing of the past.