Contact centers in 2026: trends, challenges and the role of AI

The webinar has two parts. Oksana Budz opens with her impressions from Customer Contact Week in Orlando a customer service conference running for over 25 years. In her view, most attendees were not vendors but leaders who came to learn. She singled out three challenges: context gets lost when a chatbot hands a conversation to a live agent, so the customer has to repeat everything; personalization in self-service is still weak; and poorly implemented AI becomes just another barrier for customers, much like the old basic chatbot.

At a round table of about 30 CX leaders, the question was what their strategy for 2026 and 2027 would focus on: people, technology, processes or customers. By Oksana's own count, roughly 85% chose technology. The webinar audience mostly picked people and customers. The top priorities of US leaders were conversational AI, a shift to self-service, analytics, the knowledge base and lower costs. Among technologies, agent assist tools for operators came first followed by customer analytics and chatbots. CSAT and FCR are becoming the key metrics, and NPS, which used to sit in the top three, has dropped back.

Iryna Velychko, who went through about 20 different reports to prepare, explains why she sees 2026 as a turning point. Service quality now means speed, accuracy, trust and empathy, and the contact center is becoming a basis for company decisions. She describes an approach where requests are rated not only by type but by urgency, complexity and emotional tension, to decide which channel and which person should handle them. At the same time the "zoo" of systems keeps growing, and so do errors at the points where they connect. The first step toward integration, she says, is to write down the processes.

Iryna calls quality assurance and workforce management (WFM) the nervous system of service. Modern WFM should work in real time, adding or releasing operators as the load changes. The second pillar is one knowledge base shared by agents and automation, with an owner, update rules and quality metrics. Metrics are changing too. Handle time and other basic figures show how the center runs, but they are not what decisions are based on. Management understands money, so these numbers need to be translated into money, with a clear return on every system.

Voice is not going away. Whatever companies do, it still carries more than half of all contacts. Self-service cannot replace operators, because people want to talk to a person and bots cannot close complex cases. Each center has to decide what share of contacts it can realistically automate. Omnichannel is held back by the same three things as ten years ago: integration, budget and processes. The cloud, Iryna notes, is no longer as cheap as promised, since call recordings and AI need a lot of resources. The agent profile has changed as well: emotional intelligence and adaptability matter most, and retention depends on career tracks, mentoring and training supervisors.

On AI, Iryna stays sober. Studies promise a lot, yet only 14% plan to roll it out this year, and Europe is the most cautious because of regulation. When she asked the audience who had used AI in contact center operations last year, only a handful responded, while one report claims 80% of centers. Ukraine, she believes, should implement carefully and show real cases to its European neighbors. For Ukrainians, service is now part of psychological safety, so the task is to make automation humane. She closes with the association's initiatives: Telegram working groups, industry mentoring for new managers, work against abuse of operators and the problem of fraudulent call centers. In the Q&A, listeners asked how to track the emotional load on agents.