In an era where consumers have high expectations for customer service, call centers are grappling with a variety of challenges. Peak volumes, staffing shortages, and high turnover put pressure on service quality, while a single complaint or news story can immediately impact brand perception. AI offers a solution, from faster responses to a consistent tone of voice. But how do you go about it, and where do you start? Our expert Michel Heijman explains.
Peak volumes, staffing shortages, high turnover, and inconsistencies in the brand image—these are just a few of the many challenges call centers face. And this is happening in an era where consumers have high expectations for customer service. How can AI contribute to the very best customer experience, more efficient operations, and better brand alignment?
Whether they’re large call centers that outsource or smaller contact centers, they all have to deal with demanding customers. “In many cases, as a call center, you actually come into contact primarily with people who have a complaint, are frustrated, or are otherwise dissatisfied with your service or product. That’s precisely when excellent customer service is crucial to avoid losing customers,” says Macaw’s CTO Michel Heijman.
You can’t deliver that kind of excellent customer service with long wait times caused by staff shortages or inefficient work processes, or with agents who can’t provide the right answers or speak in a manner consistent with the brand. “But this is the reality many call centers face. They struggle with high staff turnover, and scalability is also a challenge. Seasonal spikes are predictable, but other issues aren’t always. Think of that one news story about your brand that suddenly pops up, or a weather forecast that has an immediate impact on your service or product. New or temporary agents must then quickly become sufficiently skilled and aligned with your brand; agents need to be able to work quickly to help more customers.”
AI as a personal assistant—for example, in the form of an internal chatbot—offers call centers a solution to these challenges. “With an AI chatbot, as an agent, you can search through all available information with the press of a button and get an answer to your question in the right tone of voice,” Michel explains. “A major health insurance provider currently relies entirely on AI bots for its agents during seasonal peaks. This allows them to work more efficiently, reduces wait times for customers, and improves the quality of customer service: people receive satisfactory and accurate answers to their questions more often and more quickly.”
Because that’s what real-world experience shows: AI chatbots are very accurate. Michel: “No matter how well-trained your call center agents are, on average they provide the correct answer to 70 to 75 percent of customer questions. A chatbot’s accuracy is much higher: around 90 to 95 percent. Yet people tend to think: ‘So that thing isn’t reliable enough!’ That’s why the chatbots we provide at Macaw always include source citations with their answers. This way, people can check for themselves whether the information is correct if they have doubts about its reliability. And it helps you emphasize to your agents that their human touch remains valuable.”
In addition to providing reliable information, AI can also help call center agents better align with the brand. For example, a beer brewer launched its own AI chatbot on Microsoft Azure OpenAI Service (now part of Microsoft Foundry) within Microsoft Teams, giving employees secure access within the internal ecosystem. This way, call center agents always have access to internal information and immediately receive the correct tone of voice that aligns with the brand image and the business context. “Employees especially appreciate the Azure chatbot’s semantic search function: the bot takes the context of words into account, allowing it to search for information related to the customer’s query. This truly takes customer service to the next level,” says Michel.
How do you ensure that the bot—as in the case of the brewery—streamlines communication and aligns it with both the brand and the customer? Michel’s advice is to configure the chatbot’s prompts as effectively as possible on the backend. “For example, tell the bot in your technical prompt: Your name is Lisa, you’re 30 years old, you’re service-oriented and helpful, you have a warm, human tone, you use simple language, and you embody our core brand values: we are ethical, sustainable, and committed to high product quality,” Michel explains.
As a call center, are you unsure whether to start experimenting with AI now or wait until later? “Our advice is: don’t wait—now is the time. AI is here, and with the EU AI Act, the rules of the game are now clear, so you can build with governance in mind from the start. See what it can do for your call center in terms of efficiency and higher customer satisfaction. It’s also an opportunity for your employees to learn new skills that they’ll need in the future anyway,” says Michel. Your data doesn’t have to be fully organized yet, and an experiment doesn’t have to be expensive or take a long time. “In our programs, you can start your experiment internally after just 6 weeks. We’ll help you identify the benefits for your call center up front, and 6 weeks later, an AI pilot will be up and running in your organization.”
So the motto is: just get started. Standing still is not an option. How can your company turn the AI hype into value? Contact our experts right away.