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
As one of the largest distributors of workplace products, Lyreco's expertise covers workplace needs, like office supplies, furniture, IT equipment, printing solutions, personal protective equipment, hygiene, coffee and catering. Therefore, they have a huge and ever growing portfolio of products, which sometimes differ only gradually. Hence, Lyreco continuously needs many high-quality product descriptions.
The Challenge
Before using AX Semantics, Lyreco already had a lot of texts, but they didn’t meet the company’s requirements. Also, the workflow was very inconvenient and time-consuming, as the texts had to be manually transferred into an excel file. Moreover, writing 50 to 100 product texts a day and gathering all the needed information for it, was a very time-consuming and unsatisfying task for the team members. As the writing process involved many different people, the product descriptions differed widely in style, tonality, structure, and the overall text quality.
Their goals:
- Automate the content creation process as quickly as possible
- Save time and money
- Relieve employees from monotonous tasks
- Standardize high-quality product descriptions
The Solution
The Lyreco team knew they needed to find a way to automate their product descriptions. So they decided to use AX Semantics.
Structured data
In order to create automated texts, Lyreco had to tackle their data challenge and create a structured and complete data set. In creating clusters, Lyreco pooled products with the same attributes into categories and filled out all the missing data and attributes for their products. For example, for pens, paper, highlighters, permanent marker attributes can be color, brand, refillable, line thickness and so on. Additionally, Lyreco improved their taxonomy.Scalable product descriptions
Lyreco uses a PIM system to feed all their product data into the software of AX Semantics. Therefore, new texts are created very quickly with the AX Semantics software, and are instantaneously published on the website. The descriptions are scalable and the data is easily turned into high-quality texts. This automated process saves a lot of time and money, in contrast to the content that had to be written manually before.High-quality texts
Lyreco’s product descriptions are consistently of high quality. This is because texts are now created based on rules and data set by Lyreco, in contrast to the manually written texts that were created by many different copy writers. Moreover, typos, spelling errors, differing tonality and writing styles in texts have been eliminated. Also, the new product descriptions have a standardized text structure, like paragraphs or formatting, and contain all important USPs as well as relevant links. This means that no corrections or proofreading is necessary, even when Lyreco includes new products into their portfolio.
Product description for a toner on Lyreco’s Website
The Results
- Due to content automation, Lyreco is able to standardize a high quality for their product texts.
- Before using AX Semantics’ software, Lyreco needed about ten minutes to write every new product description manually. Now, the content creation is scalable and executed by the push of a button. In this way, the Lyreco team was able to create approx. 33,000 product texts in only two months and about 50,000 product descriptions in the first few months. Obviously, the use of the software saves a lot of time and money for the company.
- With the time saved, the team of Lyreco was able to add many important text modules to the product pages, like call to action buttons and environmental certificates and, therefore, improved the overall user experience.
- Also, in eliminating monotonous work, team members have much more fun and are able to tend to other important tasks.
