by Andrea Ludwig, Account Director, Cubewise North America

Who am I, and why am I writing this?  

When I started my career at Neiman Marcus over 15 years ago, we were told our job was simple: Get the right product to the right place at the right time.  

At first, this sounded so straightforward, because I had Excel and was eager to get it right. 

Fast forward to today, things have changed a lot. Gone are the days when you had two or three purchasing options; where same-month delivery was best-in-class, and the consumer could only buy what was available. With the internet, smartphones, and smarter logistics the consumer has been given the power to buy what they want, when they want, whenever they want and have it delivered that day.   

These concepts are not only retail but also apply to manufacturing, B2B, or any organization that needs goods at the right place to keep the business moving. Sitting on the wrong product type or not having the right parts to keep production going costs money. Whether one is missing retail sales, delaying the production of a widget because the raw material is missing, or there is a jet fighter out of commission because one doesn’t have the correct spare parts to fix it – this all leads to a loss in productivity.  

Stepping further into this challenge, organizations increasingly have access to so much data, which can complicate matters without the proper analytical systems and processes that drive business leaders to make informed business decisions. Excel is amazing, however it can and should only get you so far.   

Unfortunately, too many organizations manage their business in offline Excel. There is much room for error when using Excel, and there are only so many KPIs one can consider. Consider trying to weigh last year’s sales against the previous year’s sales and then factor in sell-through or gross margin. Can one person be consistent and objective when weighing all these factors in their sales plans or inventory allocations? We can take your organization out of offline Excel using IBM Planning Analytics and automate the process. Whether we do simple automation that allows one to plan sales weighted off several KPIs or more advanced techniques using AI, many options can help a company do better.  

One of the more advanced options we can explore is machine learning and AI. A company can use its historical data, whether for sales plans, expense planning, inventory allocation, or countless other items, to forecast what will happen in the future. Let’s be clear: the projected outcomes are only as good as the history. If an organization doesn’t have enough data (we recommend a minimum of 5 years) or the history is incomplete, this will not produce an accurate result. We can also incorporate external data, such as weather, economic indicators, etc., to help see trends that affect your business.   

Another option is to use decision optimization – set parameters that create restrictions that help curate the outcome based on restrictions and goals. A few examples of this would be to take into account how much backstock a store can hold, how many couches a warehouse can stock, or setting a price floor for transfers as it costs more to move the product than mark it down. There is no limit to the amount of logic/rules that can be set, but we find that setting 3-4 typically helps to provide more accurate outcomes.  

Many organizations want to use automation and AI to improve forecast accuracy and remove manual workload for their employees, but they need help figuring out how to get started. We recommend taking a step-by-step approach and thinking about your current process and what can be incorporated into your business that will be adopted and trusted.  

There is a lot to unpack and many paths an organization can take, but take a step back and think about its high-level goals and the top 2-3 problems it wants to solve. Then, let’s have a discussion of the best way to help you solve them.