How many times have we walked past a shopping aisle looking for the desired product only to find ourselves staring at an empty shelf? And to notch up the situation a bit more, the inventory management system will not raise a flag for replenishment- as there seems to be enough inventory in the backroom and on the shelves to meet demand.
This is not a supply chain apocalyptic reimagination in a dystopian future – but a common yet scary occurrence for retailers and brands alike! It is a no-win situation when products are not available for shoppers who intend to purchase it. This mismatch between the System counted inventory vs. physical inventory is called Phantom Inventory or Ghost Inventory.
Phantom inventory is a billion-dollar problem opportunity for retailers and manufacturers to reclaim their lost sales with in-store excellence.
CAUSES OF PHANTOM INVENTORY:
While there are multiple reasons for such mismatches occurring at the same time, here are a few key reasons:
• SHRINKAGE: Phantom Inventory can occur due to employee theft, Five Finger Discount (shoplifting!), and other sundry reasons NRF estimates that Retail shrink totaled $61.7 billion in 2019.
• Breakage during various stages of the product movement
• Products becoming unsellable due to damage, yet inventory management systems assuming they are available for sale.
• Data entry or Scan issues during checkout
• Products are misplaced by customers or stocked in the wrong place
• Products “lost” in the backroom and not available on time for sale
• Human errors: Many mistakes occur during merchandizing processing, system update, date entry, returns processing, online order fulfillment, etc. To err is human!
EFFECTS OF PHANTOM INVENTORY :
• As stores depend on automated reordering systems, these discrepancies between actual vs. system inventory impact the whole process and potentially lead to stockouts
• Lost Sales – Shoppers not finding their desired product will either find a substitute product or switch loyalty to a new brand or retailer. Depending on the industry, lost sales value is usually within 2%-3% of total sales.
• Inaccurate data due to Phantom Inventory will lead to poor insights. It might be related to erroneous reporting of low sales in a store for a category or a brand. Even the launch of new products will be unsuccessful if the right checks are not in place to track Phantom Inventory from the onset.
• Inaccurate demand forecasts and plans lead to problematic demand forecasts because systems may show products as in stock but unsold when they have not made it to store shelves at all.
SOLVING THE PROBLEM:
“We cannot solve our problems with the same level of thinking that created them.”
Albert Einstein
As technologists, we all believe in the strength of a holistic approach to mitigating this tricky retail conundrum. The most optimal solution would entail a mix of data available to drive insights via retail analytics and software solutions.
Both retailers and manufacturers use multiple analytics solutions backed by sophisticated ML algorithms to verify their inventory positions. They shift through almost real-time data over multiple SKU/Store/Time combinations to spot anomalies and outliers. The first step is to know if an item is missing, and the next step would be to attribute the missing item to a probable cause.
Let us look at a straightforward yet effective way to identify potential Phantom Inventory issues. Company X has one of their laundry brand ‘ACME’ selling on an average of 25 units per day in Store Y.
Analyst Joe sees that Sales have come down to Zero for the latest day(s). Based on historical sales trends (sale-through) or demand forecast, Sales should not be zero, although it is theoretically possible. There is no seasonality or data noise to account for the zero sales. 
This gap is most likely a case of phantom inventory preventing shoppers from purchasing products from Brand ‘ACME’. A simple time series would help to identify items having Zero Sales for the last ‘X’ days and having greater than Zero Store Inventory.
If left unattended, as Phantom Inventory is not easily attributable, it will create inaccurate data, which leads to poor insights related to low sales in a store or even forecasting that will be off down the line. It is also crucial to keep a keen eye on PI when a new product launch is being monitored; else, we risk losing a future Winner brand due to lousy store compliance!
Phantom Inventory Explained: Here are a few ways to identify and prevent Phantom Inventory.
- Compare on-shelf availability to perpetual inventory: Physical count of on-shelf products can significantly help- focus on high value or fast-moving categories if prioritization is needed.
- Analytics: Inventory exception report mapped to a rule-based alerting workflow to catch issues as they arise.
- Monitor the product’s touchpoints right up to the store’s aisle to identify the root cause. Ask “Why?”. And more “Why?” until a set of hypotheses are arrived at for validation. Remember, the root cause(s) might be a combination of many factors.
- Crowdsourcing / Mystery Shoppers: Incentivize shoppers to post pictures of missing products in the aisle and seek their feedback.
- AI-backed solutions – enable retailers to use emerging imaging technology, weight-sensing technology, or RFID tags and sensors to automatically track the exit of materials from a shelf. Amazon Go has paved the way for the future, perhaps!
AI helps to identify and prevent phantom inventory:
Stores are increasingly leveraging Artificial Intelligence (AI) based System for tracking shoppers’ path within the store and shelf products. For example, Amazon Go relies on hundreds of cameras and sensor fusion in each store to identify products that are taken off the shelves by shoppers. A mesh of Artificial Intelligence (AI), Machine Learning (ML), and Deep learning leveraging image recognition technology, laser sensing, etc., can accurately identify the item on the shelf and, when taken off, returned, or bought. This will, most likely, lead to a fundamental shift in how retail sales and inventory management will take place in the future.
Machines can see intelligently via ‘computer vision’ and ‘deep learning’, allowing them to recognize patterns and draw conclusions from vast datasets. If this AI-enabled System gains traction, it will bring in tremendous supply chain optimization, even cut down Checkout Lines, reduce Shrinkage, keep track of inventory accurately, and so much more.
Leading Retailers and brands realize that in-store excellence holds the key to solve Phantom Inventory problems. Having robust inventory management with strict compliance and checks & balances will bridge the physical and System inventory gap. Phantom Inventory is a multi-billion dollar problem waiting to be solved. Technology and analytics-driven insights might just clinch it!
Comments are closed.