AI in customer service works – yet many companies fail at implementation

• Reality check on AI adoption: The share of companies using AI in customer service has dropped from 95 percent to 54 percent year-over-year
• AI delivers results in day-to-day service operations: Companies reduce response times and boost customer satisfaction by an average of 19 percent through the use of this technology
• The next step for AI is stalling: Data quality, system integration, and governance, in particular, are hindering scaling efforts in customer service

Munich, July 2026: Artificial intelligence is noticeably improving response times, customer satisfaction, and efficiency in customer service. At the same time, it is becoming clear that success depends less on the technology itself than on changes to business models, processes, and structures, as well as data quality. These are the findings of the new Roland Berger study, "When the hype fades, reality hits." For the study, a total of 550 customer service executives across ten countries were surveyed between November 2025 and March 2026.

The initial euphoria surrounding AI in customer service is increasingly giving way to a more realistic view of its actual application and added value. The fact that the proportion of companies reporting the use of such applications has dropped from 95 percent to 54 percent within a year reflects not so much a retreat as a sharper understanding of what actually qualifies as AI in customer service. Where the technology is already in use, the verdict is decidedly positive: more than half of the companies report a significant or very significant impact on their service performance.

This effect is particularly evident in operational metrics. On average, AI reduces response times by 19 percent, boosts the Net Promoter Score (NPS—a standard benchmark for customer loyalty and satisfaction) by 10.7 percent, and increases process efficiency by 11.5 percent. At the same time, operating costs fall by an average of 11.7 percent. AI is thus evolving from a purely behind-the-scenes efficiency driver into a visible component of customer interaction; in 45 percent of AI-enabled service organizations, it already serves as a key customer touchpoint. Acceptance levels are also high, with 80 percent of customers reacting positively or very positively to AI-based bots.

Despite this progress, scaling remains the biggest challenge. To date, many companies have achieved only a moderate level of maturity regarding key prerequisites such as data availability, data quality, integration, and governance. Even among AI users, legacy systems, unclear responsibilities, and fragmented process landscapes are hindering a broader rollout. The central question is therefore no longer whether AI works in customer service, but whether companies are modernizing their business models decisively enough to unlock sustainable value.

"The market is entering a new phase: moving away from high-profile pilot projects toward scalable applications with clear operational accountability," says Simone Schatto of Roland Berger. "This is precisely what will determine, in the coming years, which companies succeed in simultaneously improving efficiency and the customer experience. At the same time, pressure is mounting on companies that are currently still foregoing the holistic use of AI in customer service."

The study thus highlights a clear turning point: the experimental phase is over; what counts now is robust implementation. Companies already deploying AI in their customer service operations are realizing measurable benefits in terms of quality, speed, and costs. Those that hesitate risk competitive disadvantages—not because the technology is immature, but because the organizational foundations for successful deployment are lacking. The next stage of evolution in customer service will therefore be defined less by new tools and more by the ability to rethink processes, systems, and roles.

About the study
The study is based on joint research by Roland Berger and Potloc, conducted between November 2025 and March 2026 across the USA, Brazil, Germany, Italy, France, the UK, China, Singapore, Japan, and Thailand. The 2026 survey includes more than 550 participants, primarily executives and decision-makers in customer service. Combined with the previous year's wave, the research covers more than 1,100 unique respondents. In total, the respondents represent more than 210,000 employees in customer service units across the consumer goods and retail, financial services, energy and utilities, technology, media and telecommunications, and transport sectors.