How to use AI to optimize your article master?

calendar_today November 18, 2021
Personal image
Marketing Team

Prisma Retail

  • In the following blog post, we will analyze the use of artificial intelligence to optimize product classification in the retail industry.
  • Changes in consumer behavior and omnichannel have made it difficult to manage and sort products consistently and efficiently.b> Keeping product information up to date, error-free, and with full attributes is now a significant challenge for retailers. It is common to see miscategorized products, wrong units of measure, items missing essential attributes such as brand or size, and even the same products with different descriptions.
  • Progress in Artificial Intelligence (AI) has opened up new opportunities for companies to optimize the product classification process. For example, having an up-to-date, accurate, and consistent item master optimizes inventory management, which allows the detection of consumer patterns through product attributes.
  • What should we take into consideration when optimizing an article master?

  • 1) Have clean and accurate data: internal and external data must be captured, cleaned, and analyzed to make efficient decisions.

    2) Decide Product Hierarchy: it should be specific for planning the product display and consider customer behavior.

    3) Monitor the market: the category in a store is the same as the one stocked by the competitor. And, to avoid losing sales, tracking the competitor's movements is critical.

    4) Search for competitors' prices: matching algorithms must be available to find the price of the same product in the competitor's store.
  • The solution: simplify your item management process

  • Prisma uses Artificial Intelligence algorithms in the Item Master module. Our algorithms learn from in-house and public data to suggest the correct product category and extend its specifications with attributes (e.g., brand and size).

Generating the following benefits:

  • Reduces errors and speeds up the creation and management of products (SKUs).
  • Enhances information analysis, detecting consumer patterns by attribute.
  • Allows defining differentiated strategies for each attribute, for example, having more competitive prices in larger-size products.
  • Provides accurate information to the consumer about the characteristics of a product.
  • Autocompletes product categories and attributes, avoiding errors and enriching the data.

Prisma uses Artificial Intelligence algorithms in the Item Master module. Our algorithms learn from in-house and public data to suggest the correct product category and extend its specifications with attributes (e.g., brand and size).


We use Matching Algorithms.

  • Prisma uses matching algorithms to collapse masters and expand descriptions. Many times, in diverse stores or franchises, products are uploaded under different names. If you have the EAN, it is easy to associate items, but if you don’t have them, matching algorithm can be used. In this case, a categorization is made taking into account the description. For example, the "200ml coke" is associated with a series of attributes (size, calories, ingredients...) and, thus, we expand the item master.
  • Also, it is complex to find the price of the same product in a competitor's store since it may be listed under a different name. That's why we extend the item masters with matching algorithms to capture and analyze competitors' prices and avoid errors.

Business Case

  • Today, oil companies face the great challenge of owning franchises that have heterogeneous systems and item masters. In other words, there is an immense amount of data stored and processed by each point of sale and with different technological solutions.
  • Also, the Branded Gas Stations are afraid of losing franchisees to other oil companies that offer better prices and promise to improve their profitability.
  • Eager to offer added value to franchisees, the oil company hired Prisma. The implementation journey began with several consulting-style sessions where their status quo was evaluated, and business objectives were outlined. During the meetings, the company's operations were reviewed and the interfaces required were evaluated.
  • To analyze business performance and generate intelligent suggestions, it is necessary to have a complete history of sales, costs, taxes, margins, etc.. Data was collected from different points of sale to detect anomalies, match product descriptions and enrich information to create a complete and error-free item master.
  • With Prisma, the company was able to find equivalent products from their competitors to compare prices, assortments, and trends. Also, they were able to segment items with multiple attributes, manage multiple units of measure for stock control, purchase data, and supplier relationships, and better inform the consumer about the characteristics of a product.
  • Conclusion
  • Categorizing products is the key to short- and long-term success. Today, it is essential to define differentiated strategies for each group of items to enrich the information with market data and streamline product creation and management.
  • Prisma has an API to integrate the Item Master with the customer's diverse systems, including ERPs, POS, BI, among others. Thus, thanks to AI and Machine Learning, you will achieve a continuous improvement of your category tree and an efficient operation of the item classification process.
PRISMA is a B2B SaaS platform

You may also be
interested in.

Find out how we can help you!