By Keith Whaley and Greg Arthur
Imagine you’re playing Texas hold ‘em. At the beginning of a hand, the only things you know are what cards you’re holding, and how many other players are playing the hand. Each time community cards are dealt, your expectation of your odds is updated and you bet accordingly – betting more if you think you have a stronger hand and less if you think you are not likely to win.
But if you were to play poker the way many companies manage their inventory allocations, the betting pattern would be very different. You would blindly bet all of your chips after the first round of cards are dealt, before you have accessed your odds of winning the hand. It doesn’t take a seasoned gambler to understand that this is not exactly a recipe for maximizing winnings.
Betting with Retail Inventory
Think of the hand you are dealt as an inbound shipment of inventory and the chips you bet as the units of that inventory. You and the other players are the allocators that need to make the optimal bids with your chips in order to maximize your winnings. Given you can make better decisions the more information you have, doesn’t it make sense to start by betting just a percentage of your chips until you have more information?
Why go “all in” with your inventory allocations when you can be a more strategic player and place your bets as you learn more? Allocating products or betting on inventory allocations to stores requires utilizing strategies such as product modelling, profiling and size curve management. Additionally, it makes sense to “read” the business or the hand you have been dealt before betting your inventory. Allocating forward weeks of cover against demand enables you to react to the “reads” you’re getting from actual daily/weekly demand.
Also, not all the chips are the same. So just as you have many different types of chip values, you have different items in your inventory. The complexity multiples as you realize you have to place bets for each item. To deal with this complexity, many companies resort to using generic allocation polices. In most cases these generic policies enable the inventory to get allocated and shipped, however the inventory allocations that result are often mediocre bets at best.
A better option maybe to bet your inventory allocations based on individual item strategies rather than generic one-size-fits-all policies. Enabling this type of betting strategy can be accomplished through a system with allocation functionality that is flexible and allocators that can think strategically about placing bets on their inventory.
Betting Strategy is Important
For any product a retailer is stocking, every unit that leaves a distribution center is no different than throwing chips into the pot. Each allocation represents a bet on the sales potential of an item, and those bets need to be based around a strategy that makes sense for the item. Betting too much inventory against initial allocations just to clear the distribution center can be a recipe for a short night at the poker table.
If inventory is positioned incorrectly, the result can be markdowns or costly redistribution expenses. Implementing individual item strategies for allocations can present a significant potential source of profit by having the right products at the right place at the right time.
This is where a tightly integrated planning and execution process can help. Allocators should set their initial alloction plan and stick to it. Only after the player has had the opportunity to read their hand versus the community cards should a deviation occur in their betting strategy. Similarly, a good allocator will only deviate away from their initial strategy after they understand the actual demand patterns occurring in their stores.
Betting Techniques make a difference
Knowing the proper techniques to utilize in inventory management is critical to the overall health of sales and margin for a retailer. Understanding when to use allocation or replenishment techniques can have a significant impact on your inventory ROI. A core product with a six month lifespan and a relatively stable demand level is going to require a much different strategy and technique than a four-week fast fashion item with high upside coupled with extremely high volatility. Not all items are meant to be allocated and not all items are meant to be on forecasted and replenished. Assigning strategies and techniques to items based on anticipated behavior can be an extremely profitable approach.
Learning to place good bets
At JustEnough, we recognize the immense benefit leveraging the computational potential of an allocation system can represent, but also understand the limitations of what software can realistically accomplish. We’ve engineered an allocation solution that lets each user define a strategy for each item, defining the information a machine can’t know, such as what to model a new product on, where the product needs to go and how aggressively the user wants to bet.
At the same time, we take away the need to consider your supply chain on a daily basis, and by defining an exception-driven workflow, make sure that planners can spend as much time as possible improving item-level strategies. And once a strategy has been defined and actual demand signals begin to be recorded, the forecast of each item at a location level is continually updated to ensure you are maximizing sell-through and improving profitability. Now that’s a good bet!