Welcome to our Support Center
< All Topics
Print

Step 3. Discount recommendation: Algorithm selection

Algorithm selection
Use Algorithm selection screen to select the algorithm for existing project. You can create multiple projects for the same algorithm to support multiple output requirements.
Discount Recommendation
Use Discount Recommendation algorithm to assign appropriate discount levels against respective customers and products. The output of algorithm will help you to identify what is the level of discount to be provided to the customer for a certain product in order to keep the customer interested in your product but on the other hand, not to provide a too high discount as the customer will be buying the product anyway and it might be leading to unnecessary revenue loss. This algorithm is running solely on local computer and none of the data is stored on the cloud.
Algorithm Mode
The output CSV file of this algorithm might have a big size which can lead to issues when the file is uploaded back to the ERP/CRM system and therefore user has an option to select one of the two algorithm modes. General recommendation on algorithm mode selection is obviously the type of pricing which is used in your company (per product or per product category). Another pointer might be to multiply a total number of product lines with a total number of customers. Mode “Output: One row per product/customer” is usually recommended if this number is not exceeding 1.000.000 rows however this is not restricting users to use this algorithm over 1M rows as import of outputs is strongly dependent on the methodology of importing output files and capabilities of ERP/CRM system.
Currency
The user needs to select default algorithm currency in order to provide the best result of algorithm output. All files to be uploaded for algorithm processing have to be in the same currency as the currency selected in this step (customer revenue, selling/purchase prices, etc.).
Tip: You can create multiple projects for the same algorithm but in different currencies in case you need to support multiple currencies.
Notes:
  • The algorithm is providing three levels of discounts; Recommended discounted price, Maximum shop-floor discounted price, and Management discounted price.
  • The algorithm has been trained on millions of lines of analytical sales data and output is fitting any kind of industry.
  • Artificial intelligence is used to recommend relevant discounts however users are still in full control of minimum margin over various parameters.
Previous Step 2. Discount recommendation: Project details
Next Step 4. Discount recommendation: Forecaster-> Sources – Data inputs
Table of Contents