How De Huidkliniek implement an AI that cares: Delivering human-like support at scale

About De Huidkliniek

Founded in 2003 in Amsterdam, De Huidkliniek has grown over the past 21 years into a trusted name in professional skin care, with 17 clinics located across the Netherlands. From the very beginning, their goal has been clear: to help people feel confident in their own skin.

They believe that everyone has a natural beauty and their mission is to enhance and celebrate that uniqueness. They take the time to truly listen, understand their customer’s needs, and design a tailored treatment plan with expert product advice that supports skin's long-term health.

We spoke to Zoe Prevoo, the Team Lead of Customer Care, about the challenges they’re facing, the solutions provided by Trengo, and how AI has helped them tremendously in resolving customer inquiries.

The Challenges

De Huidkliniek has been using Trengo since 2022, but Zoe Prevoo, Team Lead of Customer Care, noticed they weren’t making the most of it. She saw a chance to improve things and started introducing more of Trengo’s features to help the team work smarter.

Since the clinics are spread out across different regions in the Netherlands, the team’s main challenge is responding to customer questions as quickly as possible, while also making sure no potential leads are missed, especially outside of business hours.

Keeping in touch with long-term clients is super important too. Whether it's about appointments, issues, or general questions, the team wants to make sure every message is handled carefully and on time.

That said, each clinic runs a little differently, so a one-size-fits-all approach doesn’t really work. On top of that, the live chat was picking up a lot of vague messages, sometimes just a “Hi” during non-business hours. Without any follow-up info, it was tough to respond or reach back out. So, the team decided to scale down live chat and focus more on WhatsApp, which gave them a better way to stay connected.

They were already using quick replies, rules, and labels in Trengo, but Zoe saw room for even more improvement. That’s when they discovered Trengo’s latest feature: Trengo AI, and things really started to level up.

The Solution

"In general, we focus more on the quality of the response than on speed."

WhatsApp

They decided to focus on WhatsApp, especially for handling customer inquiries that come in during non-business hours. They also noticed that responses came in much faster through WhatsApp compared to other channels.

“Most of the time, we get a reply from our customers within the hour.”

Previously, with live chat and mandatory forms, potential leads would sometimes abandon the process and leave the page. However, with WhatsApp, no leads slip through the cracks—they already have the customer’s phone number, which makes it easier to follow up.

“We didn’t even see a drop in engagement when we deactivated live chat and shifted our focus to WhatsApp.”

They also use WhatsApp to send confirmation messages, for example, when a customer cancels an appointment or wants to reschedule.

Trengo AI

When peak season hit and with one fewer team member on board, they needed support more than ever. That’s when Alan, Trengo’s Customer Success Manager, helped them set up Trengo AI, and they haven’t looked back since.

They started with the most basic questions, like “Can I reschedule my appointment?” While customers technically can reschedule on their own, the AI helps guide them step by step, giving them the confidence and clarity to do it independently.

Over time, they added more frequently asked questions to the AI knowledge base, and the system began to learn and respond more accurately.

“…we realized we could handle our busiest season with the same, or even fewer, people on the team. I'm really glad we started using Trengo AI. I wish we had done it sooner.”

Rules and Labels

We don’t leave conversations hanging or open-ended. By setting time-based rules and labels, we get regular reminders, for example every four weeks, so nothing slips through the cracks.

If a customer cancels an appointment, the system reminds our team to follow up. We then send a WhatsApp message using a template to ask if they’d like to reschedule.

The Result

They've definitely seen a boost in their team's productivity. Their team is now free to tackle more diverse tasks and engage in more significant conversations throughout the day, rather than being tied up with repetitive customer inquiries.

Initially, their customers even mistook the AI for a human agent. It was quite amusing because they hadn't informed them they were interacting with AI.

.. but yeah you don't see the difference, the sentences of the AI are so good that you really think that you speak with another person.

The AI's responses are so well-crafted that it's genuinely hard to distinguish them from human communication. Now, they open chats by letting customers know they're talking to an AI agent.

They're keen on maintaining a personal connection with their clients, so they've trained the AI to reflect their informal communication style.

We don’t want to sound so stiff, we’re really keen on having personal connection with our clients so we teach AI to use first name, use ‘je’ instead of ‘u’. We want them to know that our clinic is for everybody so we also approach it in a more non-formal way of communication.

Their team has been excited about using the AI since day one. It was particularly eye-opening when a colleague left, as the AI became more essential than ever. It has saved the team a significant amount of time. They no longer have to wait or go back and forth with customers, as the AI has already gathered all the necessary details. It has also reduced pressure when they're overwhelmed with inquiries.

We almost see the AI assistant as an extra employee who really helps us get the work done. We even see some customers thanking the AI assistant.

Even when the AI can't provide an answer, the query is escalated to a team member, and they share that information with the customer. This makes their customers feel heard and reassured, knowing that someone will address their query the next day at 9 AM. This is a significant improvement compared to traditional chatbots, which are often rigid and lack the empathetic tone they strive for. Their AI, however, still feels very human.

For example, when a customer said, my dog is sick and I need to reschedule, our AI agent will reply, i’m so sorry that your dog is sick, when would you like to reschedule the appointment?