With the COVID-19 novel coronavirus on the rise worldwide, retailers are being forced to rethink their supply strategies on the fly. Demand for sundries and medical essentials is booming – especially via “buy online, pick up in store” (BOPIS) as well as curbside and home delivery channels that aid with social distancing. Sometimes this requires inventory relocations or fulfillment from stores outside of customers’ locales.
Therefore, retailers really need to assess where the demand truly originates. Then they can adjust their fulfillment behaviors to keep up with this trend and understand what further changes should be made as demand fluctuates in the coming months. There are several ways to successfully navigate this new challenge, but they all start with the same first step: identifying and understanding true demand and the ability to sense demand as close to real time as possible.
Determining ‘True Demand’ can be demanding without the right analytics tools
Unlike more traditional retail demand calculations, which were made by simply looking at historical data and making a prediction based on past trends, “true demand” is now based on date, time and geolocation, which is the location to which the demand should be attributed rather than consumed.
Why the change?
Physical, in-store inventories are no longer the only, or even main, metric of demand consumption. The rise of BOPIS, “buy online, return in store” (BORIS), ship to store, ship from store, direct home shipping from warehouses and even drop shipping from manufacturers has forced both brick-and-mortar and e-commerce retailers to assess inventory availability, fulfillment trends and consumption patterns in an entirely new way.
It’s not enough to know where somebody bought a given product. You also need to know where the buyer is located and where the product is being consumed. This information can be answered by leveraging geolocation — the hallmark of true demand and the key to understanding demand well enough to optimise your supply chains in the modern retail environment.
For example, imagine I’m confined at home and need to order food online from a national retailer to be shipped to my house in London. If this order was fulfilled and shipped from one of the retailer’s local London stores, the true demand location for this item is simple: it’s London, which is where the product was bought, fulfilled (i.e. from a London-area store) and consumed.
Now, say the retailer’s London store (the ideal fulfillment location) is sold out of these products and the order must now be fulfilled from a store in Manchester in order to optimise cost and shipment time.
Where should the demand be attributed, and, by extension, to where should the product’s replenishment be reallocated after the depletion of the goods? Should it go to the London store, where I live? Or should it go to the Manchester store, from where it was fulfilled? This is the dilemma that many retailers face when trying to gauge true demand and optimise their supply chains.
Assuming the retailer has an analytics solution with true-demand allocation capabilities, the retailer would find that the answer is still London — the order was fulfilled in Manchester, but London is where the demand originated and where the product will be consumed. Demand should be attributed to London, and the retailer should allocate more of the coats there to accommodate demand.
How can Prescriptive Analytics Tools help Determining True Demand?
Predictive analytics solutions have been used quite extensively in demand planning over the past decade or so. But the truth is that only prescriptive analytics can combine all these data points – and more – to accurately calculate your true demand.
In one increasingly common situation, we see prescriptive analytics used to determine the impact of gift shipping during demand planning.
For example, say I decide to buy that same food; only this time, I send it to my family in Liverpool who shouldn’t move from their home due to the current situation. Now where should the retailer attribute the demand? Should it be attributed to London, where I (the buyer) am located? To the Atlanta store that will fulfill the order? Or to the Liverpool area where my nephew (the consumer) is located (and therefore the product’s final destination)?
A less-advanced analytics solution might tell you the answer is Manchester, the fulfillment location, meaning that additional coats will be allocated there. But that doesn’t make logical sense – Manchester’s winters are usually mild, and people are unlikely to need heavy coats. You will suffer margin erosion from shipping more unneeded coats to Manchester, as well as marking down the unsold inventory at the end of the season.
A prescriptive analytics solution with true demand capabilities will identify the correct answer as Liverpool, where the consumer lives and where more food will likely be needed. In addition, the solution will automatically adjust the retailer’s allocation algorithms accordingly. This is not a capability of less-advanced analytics solutions, which neither properly assess true demand nor provide any actionability on their insights.
Actionability is crucial in supply chain situations like this. Don’t forget that demand planning is just that – planning. It does not give you, your stores or your fulfillment partners directives on what to do in the moment to adjust demand. Prescriptive analytics is the only way to effectively and efficiently optimise your global supply chain, as increasingly complex demand drives greater dependency on imports and omnichannel fulfillment and therefore more advanced model stocks.
Remember, customers don’t care how their orders are fulfilled as long as they arrive at the requested locations at the promised time, especially in a confinement situation.. But that’s exactly why advanced analytics tools create a reliable and flexible supply chain with optimised cost-to-serve models.
Guy Yehiav, Zebra Analytics, general manager and vice president, Zebra Technologies