Comparing the Rank API and the Optimizer API
The way cost computation is done is different for the Optimizer API and the Rank API.
Use the Optimizer API to further optimize the top n ship nodes provided by the Rank API. The Optimizer API accounts for real time inventory positions and node capacity consumption.
The Rank API processes a large set of ship nodes to be ranked based on SKU inventory position and cost. The Rank API computes total cost considering shipping cost, capacity cost, processing cost, stock-out avoidance cost and markdown avoidance cost for ship nodes of each item. Some of these costs are calculated offline.
Subject | Optimizer API | Rank API |
---|---|---|
Shipping cost | Cost is calculated using the rate card that is defined for carriers. Optimizer API considers all carrier services which are mapped to the ship node and can meet the 'Days to deliver' criteria. | Cost is calculated using rate card defined for carriers. Rank API considers one carrier service which has the maximum 'Carrier Mode Factor'. |
Capacity cost | Cumulative Capacity Utilization (CCU) is calculated at runtime using Processing Plan and Capacity Consumed for ship node. Cost is calculated by applying node balancing coefficients to CCU. | CCU calculation is not done at runtime. Latest CCU value for a ship node, which was calculated by Optimizer API is cached and used. Cost is calculated by applying node balancing coefficients to CCU. |
Processing cost | Cost is calculated using Processing Cost configuration data with quantity on assignment. | Cost is calculated using Processing Cost configuration data with quantity on order line. |
Stock out avoidance cost | Cost is calculated in real time for SKU using available inventory, velocity, quantity assigned, and value at a ship node. | Cost is calculated by a separate process off-line for SKU using available inventory, velocity and value at a ship node for 1 unit of SKU. Off-line cost is multiplied with requested quantity on order line. |
Markdown avoidance cost | Cost is calculated in real time for SKU using available inventory, velocity, quantity assigned, online price , markdown price, and markdown date at a ship node. | Cost is calculated by a separate process off-line for SKU using available inventory, velocity, quantity assigned, online price , markdown price, and markdown date at a ship node for 1 unit of SKU. Off-line cost is multiplied with requested quantity on order line. |
Optimizing orders with vendor (direct ship-to vendor) nodes | Handles order lines with vendor nodes. For more information, see the Optimizing orders with vendor (direct ship-to vendor) nodes section in the Request and response for Optimizer API and Idle Optimizer API topic. | Does not do any specific handling for order lines with vendor nodes. |
Where:
Value
- is the dollar amount measure of benefit when a store sale happens. It is used to evaluate the loss when a sale does not happen due to stock out. The current definition is based on sales profit. It is updated daily based on new data feeds.Velocity
- is the average daily sales velocity predicted by a statistical model. It includes walk-in and store pickups along with online sales, and is used to evaluate the probability of stockouts and markdowns. The velocity is computed daily based on the most recent transaction data feeds.Price
- is the price for walk-in and store pickup sales calculated based on daily transactions. It is updated daily based on new data feeds.Online price
- is the price for online orders calculated based on daily transactions. It is used for evaluating the worth of avoiding markdowns. It is updated daily based on new data feeds.Markdown date
- is the planned date for an SKU-node to be on markdown specified by customers and is used to evaluate the potential benefit of avoiding markdowns. It is from customer specified markdown candidate data and is updated daily.Markdown price
- is the planned price for a SKU-node to be on markdown specified by customers and is used to evaluate the potential benefit of avoiding markdowns. It is from customer specified markdown candidate data and is updated daily.