How Ferryscanner resolves 50% of queries instantly with Trengo’s AI Agent

When Europe’s leading ferry booking platform, Ferryscanner, set out to enhance its customer support, they weren’t just seeking small improvements – they wanted to revolutionise how they served their customers. We spoke with Alvertos Iskinatzis, Customer Success Manager and Savvas Delisavvas, Customer Success Agent at Ferryscanner about their motivation for adopting AI in customer service, the incredible impact it’s had on both their support team and customers, the implementation challenges they encountered, and overcame, and their vision for the future of AI-first customer support.

Ferryscanner has been a long-time Trengo customer, and they initially chose the platform for its omnichannel capabilities, helping them solve high volume queries across channels like Facebook, Instagram, live chat, and WhatsApp. Today, Trengo AI is able to complement Ferryscanner’s existing workflows, making customer interactions smoother, faster, and more efficient than ever.

Since implementing Trengo’s AI tools, Ferryscanner has reduced the escalation rate for general inquiries by 50% and got raving customer reviews highlighting quick, efficient, and seamless experiences.

Let’s dive deeper into their journey and see how they made it happen.

Can you tell us a little about Ferryscanner and your roles?

Ferryscanner is a leading ferry booking platform across various destinations in Europe and beyond. Specialising in ferry travel, the platform connects customers to routes in countries such as Greece, Italy, Spain, and Sweden, with plans to expand further.

I started as a Customer Success Agent 16 years ago and grew into my current role. As I progressed in my career, I realised that not only could I meet my customers' needs, but I could also grow as a leader. Over the years, my team expanded steadily. I was promoted to Customer Success Manager, acting as a key liaison between various departments, including finance, technical, human resources, and commercial. I currently oversee a team of 11 people, ensuring smooth collaboration and effective management across all areas. During the busy summer months, we also hire up to 35 temporary Customer Success agents. My colleague here, Savvas, is a brilliant Customer Success agent who knows the ins- and outs of customer communication.

What motivated you to explore AI solutions for customer service?

Before AI, our human agents spent a lot of time answering repetitive, general questions. Around 30-50% of inquiries were frequently asked questions. This was time-consuming and diverted our focus from complex cases. Additionally, our old systems often resulted in abandoned chats, and there was no efficient way to manage multilingual conversations.

AI has been a game-changer. Using Trengo’s AI tools, we’ve eliminated most general questions through automated responses. This has reduced our workload significantly.

The AI provides summaries of customer chats, making it easier for agents to understand customer needs quickly. It also categorises requests more accurately, removing the need for manual tagging and complicated rule-based systems.

Two AI features stand out:

1. Trengo’s AI agent (HelpMate): The HelpMate manages FAQs and general inquiries, serving as a virtual support agent and sales agent. The AI agent is trained on PDFs, internal documentation, website, and tone of voice, acting as a human-like Ferryscanner teammate. One of the standout improvements for Ferryscanner was the ability to manage multilingual support seamlessly. The AI agent already resolves the majority of queries independently, but when an issue is escalated, it intelligently ensures it’s directed to the appropriate team based on context or language.

"The AI automatically routes chats to the appropriate agents based on language, ensuring that Italian agents handle Italian queries, and so on. This has significantly boosted both efficiency and customer satisfaction."

1. AI Journeys: The Journeys collect key information - like booking IDs and email addresses -before assigning cases to agents. This means agents can focus on resolving issues rather than gathering basic details. We’ve also integrated AI to handle requests like booking cancellations and ticket changes. Customers can now make these changes themselves, which has reduced agent workload even further.

Did you face any challenges while implementing AI?

Absolutely. Initially, the AI responses felt robotic, and we were worried it might disrupt the user experience. However, with continuous feedback and training, the AI now delivers responses that feel natural and conversational.

The most important step was getting our knowledge base ready. We compiled over 70 PDF files, covering a wide range of topics such as pet travel guide, luggage policies, cancellation procedures, and guidelines on when customers should arrive before departure.

With time and a great deal of support from your team and Trengo’s customer success experts, the AI’s responses became impressively human-like. Our reviews over the last six months reflect this change – customers appreciate the seamless support.

When the web search function was introduced, things improved even further. The replies became more accurate as the AI began pulling information directly from our website, including FAQs and ferry route details from previous years – a common request from our customers.

"This has saved us a significant amount of time. We’ve reduced the assistance rate from nearly 50%, meaning that five out of ten general customer conversations are now handled entirely by AI. This has been a huge win for us."

Additionally, your team suggested a brilliant solution for filtering customer languages and assigning queries to the appropriate teams or colleagues. This has greatly streamlined our operations.

What impact have you seen since implementing AI with Trengo?

The impact has been huge. We’ve reduced the escalation rate for general inquiries by 50%. This gives us more time to focus on complex cases, like ticket changes, which are the most frequent requests we receive.

With translation features, we can now respond to customers in their native languages. This reduces misunderstandings and enhances the overall experience.

Since we rolled out Trengo AI, we’ve seen strong results in number of areas:

Resolution rate: HelpMate resolves 50% of conversations instantly without human intervention.

Response time: We’ve achieved an average 2-minute resolution time for customer inquiries.

Chat abandonment rate: We’ve reduced chat abandonment rates by 25%.

Positive reviews: Reviews mentioning quick and efficient assistance have increased.

What’s next for Ferryscanner’s customer success team?

Our future plans include integrating AI more deeply into our platform and rolling out Trengo’s AI enhancements even further. Here are some ideas we are exploring:

1. Personalised customer experience: For instance, we want the AI to provide personalised experiences for logged-in customers. Imagine a customer contacting us and being greeted with details about their current bookings – it would save time and make the process more efficient.
2. AI on email: Expanding AI to our email channel is an exciting next step for us. It will not only bring greater efficiency in handling FAQs but also ensure that complex cases are routed to the right departments, like finance, more seamlessly. This will further enhance both our internal workflows and the customer experience.
3. AI actions: We’re also working on enabling the AI to handle detailed price breakdowns and even complete booking changes directly in the chat. Our goal is to make the customer journey as seamless as possible.

What advice would you give to other companies that are hesitant about adopting AI in their customer support functions?

Our advice is simple: try it. AI allows your team to focus on more complex and creative tasks. It’s not just about efficiency; it’s about staying competitive. In a few years, AI will be a standard tool for customer support, so the sooner you adopt it, the better.