Artificial intelligence (AI) is changing the fashion industry at a rapid pace. While some companies are taking a wait-and-see approach, others are taking bold steps forward.
In this fourth part of the “AI in Fashion” series, FashionUnited speaks with Stijn de Knop, operational and financial director of the Belgian shoe company Torfs.
How does Torfs use artificial intelligence? What do you use them for?
We try to integrate AI as much as possible. Our data team consists of five people. Our first data scientists have been with us for around six years. You develop your own applications and connect to external tools that are useful for the business.
We have internally developed an AI model that predicts which goods should be placed where. It’s not just about inventory allocation. The forecast also includes an intelligent component for defining “product space”. The system learns from customer and purchasing behavior which products are being considered. If a customer has two pairs of shoes in their sights and one of them is sold out while the other is still available, we don’t need to immediately restock the sold-out pair.
We also work with other partners, such as Markmi. This is an AI-powered Markdown tool that helps set product prices during sales.
Do you have any other specific examples?
Yes. Thanks to GenAI, the use of AI has accelerated rapidly. We use Nano Banana, Google Gemini’s AI image generator, for marketing and e-commerce images. With a single photo shoot we can create up to 30 looks by digitally combining outfits and shoes.
In addition, our web texts receive support from AI. The product descriptions are mostly created automatically based on photos. ChatGPT checks, supplements and monitors articulation.
We recently developed our own chatbot for customer contact in a short space of time because a standard solution did not meet our needs.
What are the successes and challenges of your AI journey so far? Was everything an immediate success?
The use of Markmi was very informative. We learned that we sometimes discounted too much and that smaller discounts often worked just as well. This enabled us to improve our margin.
Implementing the merchandise tool to redistribute inventory proved more difficult. The tool works well and delivers measurable results. Sales of newly distributed shoes increased by up to 50 percent. However, employees have to get used to the idea that an algorithm is making decisions that they previously made themselves.
The workload also plays a role. In business it is difficult to find enough staff. The existing team understandably prefers to devote its time to customers rather than to the predominantly logistical tasks that arise from using the tool.
How does Torfs deal with this? Do you continue to invest in it?
Yes. We try to deal with this consciously and want to invest in additional support or staff to enable maximum integration. We are currently examining whether we can adjust the norm for sales per employee per hour to reduce the workload.
How do your employees otherwise feel about AI? Is it already widely accepted?
That would be putting it too strongly, but development has clearly been initiated.
We recently organized training and inspiration sessions. This allowed our employees to discover what AI programs like Copilot and GPT can do for them in the administrative area. That caused enthusiasm. After a few days, colleagues asked for licenses and additional tools. Of course, there is also a group that is a little more reserved. Integration takes time, but the most important thing is that the topic is present.
The energy of our data team is fantastic. They are real nerds, in the best sense of the word. They are intrinsically motivated to work with AI and that is contagious. It brings curiosity and enthusiasm into the company.
What would be a dream application for the future?
It’s a bit far-fetched at the moment, but in the future we would also like to use AI for customer activation. We are currently not systematically testing which content works. We dream of a system that automatically tests and learns what resonates with different customer groups. It should then make new suggestions itself. A kind of self-learning marketing.
One final piece of AI advice for others
AI is so fascinating that you just have to dive in and discover what it brings by trying it out. My advice: Don’t slow it down and don’t budget too tight.
This article was created using digital tools translated.
FashionUnited uses artificial intelligence to speed up the translation of articles and improve the end result. They help us to make FashionUnited’s international reporting quickly and comprehensively accessible to a German-speaking readership. Articles translated using AI-based tools are proofread and carefully edited by our editors before they are published. If you have any questions or comments, please email [email protected]