Allocation Optimization

As mentioned in previous sections, the Allocation calculation is divided into two core engines: Demand Forecasting and Global Optimization.

Throughout the Optimization stage, Nextail’s algorithms try to maximize sales from the unconstrained demand while considering a wide range of business constraints (logistical & merchandising). 

This process allows retailers to adjust, refine, and secure the implementation of different inventory strategies, balancing operational constraints and strategic business goals.

By leveraging multiple parameters and features available at Nextail’s platform, users can tailor allocation decisions dynamically, adapting with precision and efficiency to ever-changing market conditions.

Parameters and factors

In this section below, it is listed the key components evaluated by the Optimization engine at all times during each scenario calculation to provide the optimal solution with the highest economic contribution for the retailer business.

• Active Range (Layout Master FA and SKU-Blocks)
  Manage whether products or specific SKUs are part of the active assortment for any given store, or if they are out of range and should no longer be carried.
  
  
• Service Level (Product and Product-Warehouse Sales Threshold)
  Set the product availability target across your whole network, regardless of the Distribution Center, or set a specific one at any given facility.
  
  
• Selling probabilities
  Based on the final SKU target stock (demand-driven, service-level-driven, or business-driven), selling probabilities are calculated for each sales unit, ranging from 1 and up to the target stock, based on a statistical Poisson distribution (several million probabilities calculated in a few minutes).
  
  
• End Dates (Product lifecycle end)
  Set specific end dates when planning to withdraw whether entire collections or specific products in one or multiple locations, and let the system adjust the allocation demand and needs until that date and not beyond, saving the shipment and extended costs of sending unnecessary inventories.
  
  
• Display Criteria (Visual Rules)
  Set the Merchandising Rules that control when a product is eligible to be placed on the sales floor. The system checks these conditions for each product–store combination:

• If all conditions are met, the product remains on the shop floor and is available for sale.
• If conditions are not met, the product is flagged as ineligible for display and is assumed to be kept in the stockroom, with no sales opportunity until the criteria are fulfilled.
  
  

• SKU Min Displays
  Configure additional and granular Display Criteria for specific products and stores by determining the minimum number of units that must always be on the shop floor after each delivery, to keep shelves and displays looking full, as defined by Visual Merchandising standards.
  • Hard condition: to be prioritized, whether completely or partially, but over and regardless of the display criteria, estimated demand, warehouse stock levels, and pack sizes & structures.  
  • Soft condition: let the system evaluate if being attended, completely or partially, or not, based on the display criteria, estimated demand, warehouse stock levels, and pack sizes & structures.
    
    
• Min IAQ
  Force the allocation regardless of the estimated demand determined by the demand forecasting algorithms, until securing at least the Visual Rules in each store.
  • Criteria A, first the demand. Let the system attend to and secure the demand and Visual Rules for as many locations as possible. Only then, if there is still Warehouse stock remaining, force the allocation to those locations where there is very low or no demand until it runs out of stock or all stores are fulfilled.
  • Criteria B, the demand the latest. Instruct the system to attend to and secure the allocation to as many locations as possible, through a minimum allocation driven by Visual Rules and deprioritizing the estimated demand. Only then, if there is still Warehouse stock remaining, attend to the demand driven by selling probabilities in each SKU-PoS until it runs out of stock or all stores are fulfilled.
    
    
• Min & Max Prepacks
  Force and limit prepacks allocation regardless of the estimated demand. Secure at least a minimum number of prepacks is assigned to each location until they run out of stock, and restrict a maximum number of prepacks assignation as well. In case there is still demand to be attended, it will do so with the remaining pack size available.
  
  
• Min & Max Targets (Product level)
  Force and limit the total allocation quantity for any given product, regardless of the estimated demand and allocation needs identified by the system, and securing Visual Rules at all times.
  
  
• Logistic Sets (Packs and Prepacks)
  Let the system know the structure of the logistic bundles, evaluate how do they fit with the identified allocation needs, and determine how to be prioritized, picked & distributed.
  • Packs: composed by a single SKU and a variable number of units.
    • Single units: 1 unit per pack
    • Multiple units: multiple units per pack (2, 4, 12, 24, 60, 120, …), customized by the retailer, to optimize logistic pick & pack process efficiency but sold separately.
  • Prepacks: composed by different SKU from the same product, customized by the retailer, to optimize logistic pick & pack process efficiency, but sold separately.
    • Flatten breakdown: same number of units per SKU.
    • Ratio breakdown: different number of units per SKU, leaning to some specific sizes (smaller, central, larger).
      
      
• Demand & offer size curve conciliation
  This mechanism, if activated, balances the demand & offer size curve for any product at all times, monitoring and addressing any significant deviation between them. That means, any large deviation between the Wh stock size curve and the store’s network demand size curve, is then automatically addressed by pushing out additional units or restricting the shipment to the most valuable sales opportunities. In this way, it minimizes the risk of facing unbalanced leftovers towards the end of the product lifecycle or Wh stock availability.
  • Push effect: sizes initially purchased with a high size weight than the one estimated by the demand finally, are then pushed out to higher ratios.
  • Restriction effect: sizes initially purchased with a lower size weight than the one estimated by the demand finally, are then restricted to lower ratios.
    
    
• Warehouse scarcity (Residual Value)
  Warehouse stock is not limitless. Due to its scarce, higher or lower, a residual value is assigned to each SKU in order to give it a fair value and evaluate the benefit of sending it out or keeping it stocked. When the remaining stock has a residual value higher than the benefit of pushing it out to any particular location, these units will remain stocked regardless of having available stock to ship, until another opportunity cames out with a higher benefit (higher probability of selling or higher selling price, for instance). This scoring is conducted automatically by the Optimization algorithms, not being able to manipulate them, in order to maximize sales and minimize overstocks.
  
  
• Store prioritization and tiebreak criteria
  When there is stock scarcity at the warehouse for a given product or SKU, the system evaluates which needs have to be attended first, completely or partially, and which ones latest. Considering that there are several factors that determine the allocation needs, and in order to maximize sales and business efficiency, the system prioritizes them in the following order as well as applying the following tiebreak criteria:
  • Store Requests
    • By total store demand, from highest to lowest.
    • By total store request, from highest to lowest.
    • Partial fulfillment is allowed if the warehouse stock runs out.
  • Min prepack
    • By total store demand, from highest to lowest.
  • Hard Min Display
    • By total store demand, from highest to lowest.
    • Partial fulfillment is allowed if the warehouse stock runs out.
  • Min IAQ*
    • By total store demand, from highest to lowest.
    • *If configured to be prioritized before the demand.
  • Expected benefit (probabilistic demand * gross revenue)
    • By SKU-PoS-Unit selling probability * selling price, from highest to lowest, ranging from 1 and up to the target stock.
    • Partial fulfillment is allowed if the warehouse stock runs out.
    • Selling prices with or without deducting taxes.
    • Selling prices considering local currencies and exchange rates.
  • Min IAQ**
    • By total store demand, from highest to lowest.
    • **If configured to be prioritized after the demand.
  • Soft Min Display
    • By total store demand, from highest to lowest.
    • Partial fulfillment is allowed if the warehouse stock runs out.
      
      
• Logistic sets prioritization (Packs and Prepacks)
  When for a given product there are multiple and different logistic sets available to attend the allocation needs, prepacks will always be prioritized over any other unless shipping the prepack generates an overstock (or additional overstock) higher than a fixed ratio. Generating an overstock means sending more units than the strictly needed. This overstock ratio is set at 1:4, which means that a prepack is proposed to be sent out as long as no more than 4 units in total are generated as overstock in any of the sizes included in the prepack, regardless of the overstock unit position generated on top of the existing one (if so). Up to 4 units of overstock is allowed.
  
  
• Capacity Constraints
  Some logistic constraints are also supported for the Initial Allocation solution, where these limits will be secured and the outcome adjusted accordingly.
  • Min Order: no allocation proposed for a store if the total units to allocate, within the scenario scope, do not achieve the minimum order shipment.
  • Max Order: in case the total units to allocate for a store, within the scenario scope, exceeds the max limit, the excess of units will be automatically removed and adjusted until meeting the maximum order shipment. The removal criteria is forward-coverage-driven, removing units first from those items with the highest overstock.
  • Max Store Capacity: in case the total units to allocate for a store, within the scenario scope, plus the current stock on-hand within the whole assortment (current and past seasons, active and inactive products, leftovers and carryovers) exceeds the max limit, the excess of units will be automatically removed and adjusted until meeting the maximum store capacity. The removal criteria is forward-coverage-driven, removing units first from those items with the highest overstock.
    
    
• Currencies & exchange rates
  Selling prices are supported with local and foreign currencies. That means, for conducting a meritocratic and fair distribution that helps to maximize retailer’s sales, the Optimization engine evaluates the economical contribution of each location under the same baseline, swapping foreign prices and currencies into a single currency baseline. This currency baseline is the one defined by the retailer for the whole platform, and local prices and currencies are automatically provided by the retailer in each daily data integration.
  
  
• Linked Lines
  In case a product has been purchased with multi-sourcing, which means with different suppliers or manufacturers or fabrics or materials, but it is sold as the same product with no differences for the consumer, the master data from all these secondary product references are switched into a single chain led by the active or parent reference, and treated as a single product in terms of calculations and reports. When having to pick & pack the product, this chain is then disaggregated by each secondary product reference individually.

As you can see, the Initial Allocation solution comes fully-equipped with all the levers and parameters above. This way Nextail helps retailers and users manage their inventory in the most efficient way, making data-driven decisions, optimizing stock flows, and aligning inventory strategies with tailored business objectives.