Eliminate Out-of-Stock Incidents With a Machine Learning-based Solution
Mahesh Athalye, Sr. Director | Client Partner at UST
It’s a worst-case scenario for any supermarket: a customer walks into the store armed with his or her groceries list but leaves disappointed as many of the items he or she wants to buy are out of stock.
Mahesh Athalye, Sr. Director | Client Partner at UST
Mahesh Athalye is a Senior Director within the retail vertical at UST. As a business and go-to-market leader, he partners with clients to drive profitable business growth by implementing retail technology – including online, in-store IT, and AI/ML solutions. Mahesh has 27 years of experience in manufacturing, financial services, and the retail IT and security sector. Prior to joining UST, Mahesh was part of the leadership team of Xpanxion (a UST Company). He played a key role in the business growth at Infosys and was part of the delivery leadership at Syntel Inc. Mahesh is actively involved in community service, and is a Board member of the TAG FinTech Society and a volunteer at the Global Thrombosis Forum.
While consumers may forgive their preferred stores for running out of essentials during this unprecedented time, they will place those with constant stocks in higher regard now and in the future. Grocery chains can minimize Out of Stock (OOS) with the right technology. Doing so will positively impact retailers’ profits and enhance shoppers’ loyalties.
Now let us examine the key factors contributing to OOS:
Lack of operational compliance (or “human failure”) at the store level
This can be created by several issues, including sub-optimal Balance on Hand (BOH), negative inventory, receiving and updating process issues, manual intervention in system suggested orders, direct store delivery (DSD) process gaps, manual purchase orders, non-compliance with planogram due to store remodeling, wrong safety stock, wrong minimum display quantity, and many other such distortions in key parameters.
Lack of coordination between retailers and suppliers at both strategic and operational levels
Many retailers have implemented forecasting systems for communicating aggregate demand. But forecasts do not include real-world input from daily sales systems, often cannot account for rapidly changing consumer preferences and needs, and rarely factor in SKUs under promotion, which typically constitutes around 40% of a supermarket’s monthly sales. In addition, retailers have limited visibility to the incoming inventory from a supplier against a specific purchase order until the supplier’s truck arrives at their distribution center. At a discovery session, they will likely identify a mismatch between line fill and case fill.
Inefficient set of algorithms built many years ago on packaged software
Retailers historically have defaulted to out-of-the-box implementations, whereas customization would take into consideration unique factors specific to their clientele, locations, regional preferences, etc. Some are from the days where “brick & mortar” was the only form of supermarket retail. Naturally, their algorithms fail to reflect the current Omnichannel reality. Since version upgrades are costly, the retailer’s IT team often builds, at best, custom enhancements complementing the existing packaged software.
The absence of machine learning input in static algorithms
It’s true that incorporating machine learning into your tech stack can have a high initial price tag, to which leaders react by saying, “One more IT system which will cost us X million and take Z months to implement and suck away precious operational /domain time.” But interpreting unstructured data through Machine Learning is the only way of consuming this treasure trove of information. The intelligence generated thereby would help formulate the right algorithm, which would then generate near-perfect Automatic Replenishment order quantity for SKU X for Store Y.
What this moment needs is the UST On-Shelf Availability Improvement Solution (OSA+), using Machine Learning Store-SKU Replenishment. We will discuss how retailers’ failure to upgrade IT services to OSA+ today and in the future can negatively impact the bottom-line.
OSA+ is tailor-made for smaller grocery chains ($1B-$10B) that may not have the resources to build machine learning algorithms from scratch and employ a team of data scientists to continually manage a homebrew solution.
OSA+ provides a finely tuned machine learning algorithm combined with a managed service delivery model. With OSA+ algorithms and our managed staff handling manual intervention and overall management, companies eliminate the significant expense of internal data scientists and stock management staff.
This solution is business-critical for any retailer with fast-cycle replenishment needs, e.g., has the same SKUs coming into their distribution center that they need to get out to stores and shelves within each store. Best yet, it can complement and enhance any existing central planning tool already deployed.
Ensuring that you have stocked up your stores is not only valuable for when a consumer is in-store, making purchase decisions but also serves as an ongoing marketing opportunity - people will frequent the stores that provide the most dependable stock.
Brand loyalty continues even during a pandemic, and grocery chains cannot assume their average customers will engage in brand substitution. Studies have shown that over 50% of people would not purchase a substitute item at a store when their preferred brand is out of stock, a real detriment to customer experience.
If a chain with an active assortment of 50,000 SKUs is tracking to a metric of 5% OOS, that means they are not able to accurately replenish around 2,500 SKUs at any point in time. Even if only 10% of those 2,500 OOS are actively impacting the core basket for an average daily customer, this will create a massive negative perception about the store in customers’ minds.
Leaving aside for a second the trust erosion leading core customers to eventually shop at places where OOS problems are less common and severe, let’s examine how those OOS negatively affect the bottom line.
Imagine a 500-supermarket retail chain generating an annual revenue of $10 billion. If the OOS problem leads to a loss of sales of even just 3% of revenue, that is a staggering $300 million. With the average gross margin rate of 20%, including rebates and discounts, that means $60 million in missed Gross Margin due to its OOS situation.
Let us see what this loss of profit means in the context of the IT budget of this retailer. With average IT spending in Retail between 0.75 % to 1.5 % of revenue, we can estimate the retailer we described above spends approximately $100 million annually towards technology, e.g., hardware, software, workforce, etc. Thus, losing $60 million worth of profit lost in this one critical area represents 60% of their annual IT expenditure!
No retailer wants its customers to leave the store with unchecked items on their list, especially when that can mean hundreds of millions of dollars in lost revenue. Deploying an actively managed OSA+ solution is the best way to fulfill the promise made to those customers: having the right product on the right shelf at the right time to meet their needs.
To learn more about how OSA+ can improve the customer experience and add to your bottom line, click here.