Optimized inventory is a must in an omnichannel landscape. By leveraging real-time modeling to uncover customer patterns, retailers can adjust assortments — and amounts — of merchandise based on when customers are ready to make a purchase, according to Kevin Stadler, president and CEO of 4R Systems.
Chain Store Age recently spoke with Stadler regarding how forecasting, risk assessment, machine learning and predictive analytics are paramount in optimizing inventory and positioning retailers for maximum profitability.
What is profit optimization of inventory?
Most retailers today realize that inventory can be an asset (if it’s in the right place at the right time) or a liability (if it’s in the wrong place at the wrong time). Through a holistic model that combines forecasting, risk assessment, machine learning and predictive analytics, we can provide a future look at inventory productivity by item, by demand location and by fulfillment location to optimize inventory for maximum profitability.
Why is this important for retailers?
With the shifting dynamics of online versus in-store behaviors, it’s clear that inefficient retailers are closing locations and going bankrupt at an alarming rate. Wall Street is rewarding both growth and profitability, and punishing retailers that have neither of those. Today, in peer retailer investment, the leaders are getting rewarded with investment, so you must stay ahead of your peers to survive.
How has the game changed as the retail landscape becomes more digitally influenced?
Patterns shift faster than they used to. Consumers move in “packs of influence” today and change direction fluidly. Older methods of target service level availability and slow supply chains are becoming antiquated because they are expensive and slow to respond to a digitally influenced society.
What struggles do retailers have on the road to profitability?
The older perspective of inventory was very siloed and did not have a unified view that could provide profit and risk perspectives. It used to be that you would make a forecast, set a fairly high service level and then look at item profitability. An integrated model uses all of those factors simultaneously to goal seek rather than use a trial-and-error process that uses large staffs and a top-down category management approach. Today’s advanced retailers are going bottom up with big data and finding patterns using machine learning to quickly adjust on a very granular basis.
What tools do retailers need to solve these issues?
Having a real-time modeling environment that correctly weighs all factors is important. This takes a lot of data and a large, fast data structure. Then you need algorithms that can weigh risk and optimize against a number of variables all at once. To visualize all of this requires some data science and multidimensional techniques. As an example we use the Markowitz modeling formula, which is based on mean-variance analysis, to provide multiple variables to be integrated then visualized via risk curves.
What role do predictive analytics and machine learning play?
A common question we ask is “Do you remember what you had for lunch two weeks ago on Thursday?” The answer is that almost no one does.
To optimize the retail environment of thousands of items over many stores, and over a number of years, you have to use machine learning to find patterns. Then advanced predictive analytic algorithms reveal patterns to ensure you have the right product in the right place at the right time. To move to this proactive prescriptive technique, you have to have both machine learning and predictive analytics.
What results are possible?
The results vary widely from several percent improvement in revenue to double digits. Existing staffs can be more proactive and strategic. Profit improvement, of course, depends on the margins and turns of the product categories, but are significant as there is both an improvement in margin and turns at the same time.
How can 4R Systems help retailers in this journey?
4R has been a partner to retailers driving sales and profits for many years. We have a unique cloud optimization environment that uses modeling, predictive analytics, machine learning and goal seeking against very large data sets.
In addition, we can use our techniques to build the business case for change up front, not after you have invested. We reduce risk and maximize return for our partners at each step of the process.
Originally published on https://www.chainstoreage.com/