Back to cases Service assistance service assistance

How achieved a 23% month-to-month decrease in live chats, whilst providing a better experience to their customers.

Situation is an online platform where consumers can take out a (group)subscription to a large selection of magazines and easily read them online. Taking out a subscription, changing data and gathering information about new journals are examples of service questions that are asked to How can ensure that consumers can ask their questions 24/7/365 to enjoy their subscription?


A development in the market is that consumer expectations are rising exponentially in the field of e-commerce customer service. That is why it is important that the experience with feels easy, personal, and fast. Just like a package is delivered within a day, we also expect an immediate answer to a question. has the ambition to be available 24/7 and to always provide a good customer experience when someone has questions. Logically, due to high costs, it is not feasible to linearly scale the contact center with more live agents. How do we ensure better availability, without sacrificing the personal experience?

Our Google-powered AI assistance answers 62% of the most frequently asked questions. By means of our smart technology we can transfer the call, when available and necessary, to a live agent within the contact center. When such a transfer occurs, the live agent can view the user’s conversation history to help keep the conversation flowing and avoid repeating questions. This way you can scale up ecommerce customer service with human-like conversations. Since the introduction of the service assistant, there are no less than 23% fewer live chats per month.

This efficient form of automation creates a win-win situation for both consumers and organizations. The consumer always receives a lightning-fast response to a simple question; with a complex question, the consumer gets more attention from a live agent. By organizing the customer service in this way, the live agents of have 2 minutes more time to carefully help the consumer with questions of higher complexity.


To find out which topics could be supported via the chatbot, we did extensive research into the frequently asked questions. For this, we consulted the live chat documentation of Some of the questions turned out to be immediately suitable to be automated within the chatbot. In addition, we found that a number of questions required additional input from the user, in order to subsequently arrive at a suitable advice or answer. As a result of the research, an efficient routing was set up based on categories, resulting in smart FAQ flows per question.

An example of this is the customer question: “How do I reinstall the app?”. The service assistant of Tijdschrift then asks: “Do you have a Google (Android) or Apple (iOS) device?”. An appropriate answer will be given based on this. If someone directly indicates in the customer question to use an iPhone, for example, this question is not asked because the operating system is already known. So easy! By asking questions in smart FAQ flows, we can provide a personal answer to solve the problem.

In order to constantly optimize the AI ​​assistant, the collected user data from the customized dashboard is used. Firstly, the dashboard provides insight into which questions are most frequently asked by consumers. New questions that have not yet been answered can then be added to the AI ​​assistant on a monthly basis. Secondly, how a flow progresses is monitored, in order to subsequently make an adjustment in recurring situations. For example, we saw problems arise when activating a subscription. After implementing optimizations in the chatbot, we saw an improvement of no less than 27% in the handled rate of the FAQs.

Our results

FAQ handled rate


Drop in month-to-month live chats

2 min.

More time per live chat

Big ball graphic
Small ball graphic