Stockout Sentinel

Increase Sales by Predicting and Eliminating Stockouts

Stock-outs hurt retailers’ revenue and consumer experience, particularly when dealing with unknown stock-outs caused by ghost inventory. Previously retailers may have opted to overstocking as a solution to improve on-shelf-availability resulting in higher inventory carrying costs and space requirements.

Phantom Inventory, also known as Ghost Inventory, is an occurrence of unknown stockout due to inaccurate inventory information in operational systems. Phantom inventory can "cost" 1%-3% of revenue depending on the retailer. These unknown or virtual stockouts are detrimental for fast moving consumer goods retailers who have brick-and-mortar stores and/or offer buy-online-pick-in-store (BOPIS).

Systems showing inventory while it physically doesn’t exist in store, causes revenue impact, poor customer experience and inaccurate forecasting. There are several approaches to identifying phantom inventory issue. Stores can increase inventory cycle frequency and scope, implement technology such as RFID tags or they can deploy machine learning models.

Ideal solution for larger FMCG retailers is AI and machine learning models due to their immense dataset making it difficult to manage in traditional analytics approaches. Ensure that your stores have good On the Shelf Availability (OSA) to maximize revenue.
Get the White Paper on Phantom Inventory

Phantom Inventory Impact

There are three major areas of impact due to stockouts caused by phantom inventory:

Impact on Revenue

When products are not on the shelf they cannot be bought. Simple as that. Phantom inventory causes on average 4% loss in revenue and in some cases even more.

Customer Experience

On average a consumer facing phantom inventory driven stockouts spends 20% more time taking valuable staff focus as well. Ifthe retailer has buy-online-pick/deliver-in-store model, having inaccurate inventory information can lead toexpensive substitutions.

Cost of Inventory

When inventory is misplaced, it carries cost and cannot be sold. Investigating phantom inventory root cause can help discovery store operations issues.

$129.5Bn
Annual revenue loss in North America due to stockouts
7-10%
Average stockout rate in Europe
$1.1Tn
Global revenue loss due to stockouts
8%
Avg. stockout rate amounting
to 4% loss in revenue in the USA

Predicting Stockouts

Machine Learning Solution

Kloud9 Artificial Inventory Intelligence has been developed with leading retailers to help organizations identify location/SKU combinations that are candidates for stock-outs. Machine learning models are optimal for resolving this challenge due to large data volumes with high number of SKU/location pairs.

Predictions

Our models predict both instock as well as stockouts leveraging historical sales data. The goal of the model is to be as accurate as possible predicting stockouts for the distribution center fulfillment cycle.

Business Outcomes

Retailers can drive business benefit from addressing the predicted stockouts and instructing fulfillment to deliver additional inventory and start root cause process.

Model Performance*

96.6%

Recall
What portion of actual positivesare correctly predicted. Also known as the true positive rate (TPR), isthe percentage of data samples that a machine learning model correctly identifies as belonging to a class of interest—the “positive class”—out of the total samples for that class

81.3%

F1 Score
Balance between Recall &Precision (Harmonic Mean). F1 combines the precision and recall scores of a model. The accuracy metric computes how many times a model made a correct prediction across the entire dataset.

88.4%

Accuracy
Proportion of correct predictions in our Model. AI accuracy is the percentage of correct classifications that a trained machine learning model achieves, i.e., the number of correct predictions divided by the total number of predictions across all classes. It is often abbreviated as ACC

70.2%

Precision
Precision is the indicator of our machine learning model's performance – the quality of a positive prediction made by the model. Precision refers to the number of true positives divided by the total number of positive predictions (i.e., the number of true positives plus the number of false positives).
* Exact results will depend on quality of dataset for learning and production

Business Outcomes

Increased Revenue & Customer Experience
Improved Inventory Accuracy for Fulfillment
Develop Action Plans to Identify Root Causes
Increased Accuracy for BOPIS

Ready to Address Inventory Opportunities?

Get in touch with one of our specialists for assessment
Get in Touch!
-->