AI-powered product classification for Beslist.
Sector
(Online) Marketplace
Products
36,000,000+
Employees
140+
Beslist.nl is a Dutch website that allows visitors to compare consumer products from over 36 million items across 10,000 shops. However, with such an immense product range, the brand faced challenges in efficiently categorizing products. Manually classifying these products would have been too time-consuming and costly. To address this, Beslist.nl collaborated with Squadra, who developed a Machine Learning solution to automate the classification process. As a result, the website’s product categorization became more efficient, and Squadra continues to refine the algorithms to achieve even higher accuracy,above 95%.
Background
The challenge.
-
Overwhelming product volume
Beslist.nl provides over 36 million products from 10,000 shops, creating an immense amount of data that is difficult to manage and categorize effectively.
-
Inaccurate product categorization
Due to the sheer volume, products are often not correctly categorized into the right (sub)categories, which complicates the filtering process and hinders a seamless customer experience.
-
Inefficiency and high costs of manual classification
Manually categorizing 36 million products would require an enormous amount of time and labor, making it an impractical and expensive solution for the company. Beslist.nl sought automation to solve this issue efficiently.
Our solution.
-
Leveraging Machine Learning for efficient classification
Squadra Machine Learning Company utilized Artificial Intelligence to develop a solution for Beslist.nl's product categorization challenges. They created subcategories for existing categories, allowing customers to filter products more effectively.
-
Machine Learning to predict categories
Squadra implemented Machine Learning, a technique that enables computers to learn and recognize patterns without explicit programming. This allowed the system to predict and assign products to the correct categories based on large datasets, automating the classification process.
-
Proven accuracy and ongoing improvement
A proof-of-concept showed an accuracy of 89-94%, confirming the effectiveness of the solution. However, further refinement is underway to improve classification accuracy, particularly in categories like electronics and household products, with Squadra’s team continuing to enhance the algorithms for even better results.
The result
PowerSuite has empowered Beslist to...
Improve user experience
Automated categorization enhanced search and filtering, making it easier for customers to navigate 36 million products.
Boost product discoverability
More accurate categorization boosted the visibility of niche products, helping customers find overlooked items.
