Case Study   

MediaMarktSaturn Automates Product Descriptions

mediamarktsaturn logo


Over 300,000 items are constantly available in the online shops of MediaMarkt and Saturn. What this means for the customer is an enormous selection, but for the employees of the online team at MediaMarktSaturn, it primarily means one thing: writing, writing, and writing. After all, each product requires a comprehensive product description, only in this way can customers make an informed purchase decision.

Download now

Illustration of headphones on a black background
Photo by Franziska Knabel - MediaMarktSaturn

„Automatic text generation offers a wide range of benefits: The detailed information about the products and the way they are categorised help customers make informed purchasing decisions. The product descriptions also increase the relevance of our products for search engines.“

Franziska Knabel

Project Lead
MediaMarktSaturn Germany

Natural Language Generation in Action

The Challenge

Product information is based on the manufacturer's details - however, these are mostly just technical data and not written content, which results in inconsistent representation of products in the online appearance. Furthermore, MediaMarktSaturn was facing another challenge common to all major online retailers: the more products a shop carries, the more complex the editorial creation of unique product descriptions becomes.

The Solution

MediaMarktSaturn opted for an innovative approach, fundamentally changing the process of product description creation. Since 2019, the company's online stores have been relying on automatic text generation using Retresco's rtr textengine software. So far, the has generated over 300,000 product texts for MediaMarktSaturn based on more than 200 text models.

The Technology

The creation of product texts gains significantly in efficiency through the use of software. Natural Language Generation creates content in a matter of seconds, negating the need for lengthy coordination loops between internal actors. At the same time, NLG produces varied content – no product description is the same as another – a factor that is pivotal for the positioning of offers in search engines.

An Excellent Solution

  • ...for the creation of over 300,000 product descriptions based on more than 200 text models

  • reta Award 2021 in the category Best AI & Robotics Application for the automated creation of product descriptions

  • Supplier Retail Award 2021 in the category Best AI & Robotics Application as technology and implementation partner

MediaMarktSaturn Case Study

Download now

Your contact

Matthias Fertig

Back to the news overview