Recognizing products

In order to properly catalog competing products which allows for comparing prices between stores, you need a way to properly recognize the offered product, its manufacturer, name, and key parameters

In this industry, we distinguish between several groups of products. We have designed over 30 algorithms that analyze the features of each product. In case of hair dye, we recognize the manufacturer, series, color number, in case of perfume we recognize the brand, series, type, and capacity of the container.

“20% free” or “buy 5 get 1 free” promotions

Many products in this field are offered in promotional containers. When one store offers a shaver in 5 piece per pack variety, another one may have the buy 5 get 1 free option on sale. Should these offers be included in reports? It’s up to our clients to decide and our algorithms and our Data Quality Management team make sure it all works as ordered.

Capacity conversion

In case of cosmetics and perfume, it is very important to recognize the correct container capacity. Our algorithms can recognize that “0.1L” is the same thing as “100ml” or “100 ml”.

Foreign names

This is one of the few categories where one product can be called completely different names in different stores.

Names of scents and colors are often provided in Polish in some stores, and in English in others. The PricePanorama Data Quality Management Team makes sure that our algorithms are supplied with proper dictionaries and can catalog these offers correctly.

Monitoring of Allegro and Ceneo

The industry is dominated by very strong offline sellers, such as supermarkets and beauty supply store networks. Most of them are also active online, but they usually aren’t present on Allegro, and often also on Ceneo. While monitoring Allegro it is very important to recognize sample products from full-value products.