“With AI, we save time and build a smarter customer experience”
Sagar Schoneveld (Head of Product at Conversed.ai) talks with Fransje Nelissen (Consumer Care Lead at Marie-Stella-Maris) about the use of AI in customer service, training a digital colleague, and how AI can contribute to both more efficient support and increased conversion.
At Marie-Stella-Maris It's about more than beautiful products. The Dutch brand has been building a strong lifestyle concept since 2011, where personal care products and home fragrances go hand in hand with a social mission: with every purchase, you contribute to clean water projects.
Behind the scenes, the consumer care department is also growing. And it is precisely there that AI is now playing an increasingly important role.
We spoke to Fransje Nelissen about the development of AI assistants. Marie, her experiences training the assistant and collaborating with Conversed.ai.
“We primarily wanted to save time and organize support more efficiently.”
Fransje has been working at Marie-Stella-Maris for about 2.5 years and is responsible for all B2C customer touchpoints. This ranges from webshop inquiries and product questions to boutique-related support and general service inquiries.
In addition, she is involved in process improvement and further building the consumer care department.
“We are growing fast, also internationally,” says Fransje. “Because of that, you notice that customer service is increasingly becoming a fully-fledged department and not just something that happens on the side.”
The reason for starting with AI was therefore very concrete.
“We were primarily hoping to save time on recurring questions,” she says. “And additionally, we also saw opportunities to ultimately get more out of customer contact, for example, in terms of conversion and product advice.”
From Simple Chatbot to Real AI Assistant
When Fransje first encountered AI, she had a different image of it than she does now.
“I was actually thinking more of a kind of FAQ chatbot,” she says. “So, a system where you put in one question and get one fixed answer. Not really something that understands context or thinks along.”
That image quickly changed when Marie-Stella-Maris began developing their own assistant together with Conversed.ai: Marie.
“That's when I really noticed how much is possible and what goes on behind an AI assistant,” Fransje says. “As a user, you often only see the end result, but not how much training, data, and optimization is needed for it.”
Training Marie felt like raising a digital colleague
For Fransje, a large part of the value was precisely in that training process. Together with Conversed.ai, she brought together the first data sources: frequently asked questions, product information, existing email templates, and other support content.
“Then, for me, the most fun part really began,” she says. “Training Marie was very educational for me. You really get insight into how AI works and how it learns from what you add.”
She describes that process with a smile as raising a child.
“In the beginning, it's really like a toddler that you constantly have to explain how things work. Then it becomes a preschooler, and now it feels more like a teenager,” she says, laughing. “Overall, it's going very well, but sometimes she still has her own opinion for a moment.”
Precisely because she can actively train and adjust, she also sees how the assistant gets smarter step by step.
“At one point, I really saw that things I had learned earlier came back later in conversations. Then I thought: yes, it really works.”
An important lesson: make sure your own data is in order first.
If Fransje were to give one piece of advice to other e-commerce companies wanting to start with AI, it would be to ensure their basic data is in order.
“We sometimes ran into issues where not all product information was fully available or in the right place,” she explains. “That naturally creates gaps in what AI can answer.”
Her main tip is therefore simple:
“Make sure your own data is in order first before you begin. Then you can build and train much faster.”
In addition, she emphasizes that it helps to make someone truly responsible for the development.
“If you have a team, it's ideal to have one or two people dedicated to this. Then you can go live faster and also make progress more quickly.”
AI as support for customer service — and in the future also for sales
At the moment, Marie-Stella-Maris’ AI assistant mainly helps with first-line support and general customer inquiries. But Fransje also sees clear opportunities for the future.
“For now, the focus is really on customer service,” she says. “But ultimately, it would be great if Marie could also help with product advice and conversion.”
She thinks beyond just textual answers.
“In an ideal scenario, a customer could, for example, open an order, click on a product, and be immediately helped with damage or a defect. Or they could upload a photo, after which the AI could assess what is wrong and pre-prepare a replacement order.”
She also sees opportunities commercially.
“It would be great if Marie could also advise in the future, add products to a shopping cart, and help customers with their choices.”
A collaboration that feels personal and involved
Fransje has spoken very positively about the collaboration with Conversed.ai.
“What I really like is that the contact is personal and accessible,” she says. “Not very formal, but rather quick to adapt, easy to text or call if something is needed.”
According to her, that's where the real strength of the collaboration lies.
“I had quite a few questions at the beginning and was also critical,” she says. “But I really feel like they listened well to that and that you always thought along with us from what we needed.”
It was precisely that combination of involvement and customization that made the difference for her.
“It doesn't feel like a standard supplier-customer relationship, but much more like collaborating. You really look at what works for each customer, and you can tell.”
The size of the team also plays a positive role in this, according to her.
“You can tell there's a lot of passion in what you do. That's why it doesn't feel generic, but truly tailored to us and to Marie-Stella-Maris.”
Building a smarter customer experience together
For Fransje, it's clear that AI is not just about working more efficiently, but also about a better customer experience.
“If you handle recurring questions more intelligently, you'll have more time for the things your team truly adds value to,” she says. “And at the same time, AI allows you to increasingly meet customer needs.”
With Marie's further development, Marie-Stella-Maris is therefore looking forward to a combination of service, product advice, and smart automation.
“For me, AI is primarily interesting when it helps us assist customers faster, better, and more personally,” Fransje summarizes. “And we are really making progress there now.”
