Machine-generated text creation is a relief for copywriters and content managers in e-commerce. Anyone involved in writing texts benefits from it. There are two key technologies: generative AI and rule-based data-to-text systems. Most have already had their first experiences with generative AI, most notably GPT. But how do these two technologies differ in terms of […]
The Case
MISUMI Europe GmbH stocks over 20 million heterogeneous and configurable products in 4 languages. Despite this diversity, the manufacturer and distributor of industrial components strives to create a certain uniformity in its marketplace in order to improve user-friendliness. For example, customers experience great added value when they immediately see on the product pages that the criteria they are looking for are included in the configurable dimensions and characteristics of the product. Or by displaying the selection options for the core features on the category page, without having to open the product detail page. To achieve this, MISUMI requires standardized product descriptions with standardized terminology and correspondingly adapted headings for the configured products.
The Challenge
MISUMI manages a large amount of product information that cannot be managed manually or using generic AI. Given the enormous variety and variability of products, it is a challenge to efficiently organise the SEO and user experience aspects of the website. Standardized or detailed product information cannot be created using manual processes. MISUMI also struggles with a high level of data inconsistency due to the fact that over 700 partner brands are currently part of the MISUMI Europa GmbH product range.Before using AX Semantics, MISUMI tried to use Excel lists to provide the configurable products with an individual product description with the possible configuration areas. This helps MISUMI's customers to recognize the configured features in the product and to be sure that they are ordering the right product.
Their goals:
- Significantly reduce the effort and error-proneness of manual work for repetitive text creation.
- Respond more flexibly and quickly to market changes and adapt product information.
- Contribute to better search engine results with standardized product information that conforms to a pre-defined framework.
- Improve the user experience by providing relevant information about the product in a way that ultimately has a positive impact on conversion.
The Solution
The error-prone and impractical use of Excel for creating product information, especially for the individual description of configurable products, led MISUMI to look for an automated solution.
Automatic creation and updating of product descriptions
A product group can contain more than 4,000 items. By linking the product data with AX Semantics, MISUMI can automatically create or adapt the required product descriptions in the desired structure and tonality, and automatically highlight the distinctive features of the products.Scaling product communication
With AX Semantics, MISUMI has created a scalable framework that allows text to be created and updated at the push of a button. The automated text creation and update workflow enables MISUMI to quickly respond to market changes and incorporate them into the product information.Consistency in all product information thanks to framework specifications in AX Semantics
Targeted framework specifications, defined in AX Semantics prior to text production, create consistency in product information. The new texts and the use of relevant keywords add context to the product pages, which helps with search engine visibility.
The Results
Thanks to automation, MISUMI was able to create several thousand product descriptions in just five months. MISUMI continues to expand the automation workflow, which significantly increases the scaling effect over time. From a customer experience perspective, MISUMI is seeing increased customer interaction on the product pages.The experience with AX Semantics so far has inspired the MISUMI team to consider developing further levels of automated text creation, in addition to the development of additional categories. This could include the implementation of an automated workflow for multiple languages.
