Customer service belongs on the leadership agenda

Responsibility for the customer is often divided across the business. Sales manages acquisition, operations manages delivery, and the points between them may have no clear owner. This fragmentation matters because service quality has a direct commercial impact. McKinsey research cited in the study finds that organizations increasing customer satisfaction by at least 20% can raise cross-sell rates by 15% to 25%, increase share of wallet by 5% to 10%, and improve customer engagement by 20% to 30%. Customer service should therefore be managed as a growth issue, not only as an operational function.

AI adoption is growing, but value is not automatic

The study confirms that AI is already part of customer-facing work. In the Harvard Business Review Analytic Services survey, 42% of respondents said their organizations use AI in customer service and support, while another 42% reported its use in marketing. These were the two most frequently cited business contexts. Deployment alone, however, does not create value. If data is inconsistent, ownership is unclear, and systems use different definitions, AI scales those weaknesses along with the process.

Start with frequent, simple, and predictable experiences

The study recommends applying AI first where the experience is frequent, simple, and predictable. Narrow use cases such as routing, summarization, prediction, and workflow optimization can shorten resolution times and help teams use their capacity more effectively. This is a practical starting point because the objective is not to automate every interaction. AI should support human expertise and free service teams to focus on complex situations where judgment, context, and empathy remain essential.

Data and governance must come before scale

Customer data is commonly spread across CRM, ticketing, marketing, product, and operational systems. Each system may hold a different version of the same customer story. AI cannot resolve that inconsistency on its own; it can spread it faster. Companies need shared definitions, clear data ownership, and agreed rules for how information moves between systems. This is the real preparation for AI-enabled service. Before scaling the technology, management must decide what reliable customer data means and who is accountable for it.

Measure customer behavior, not only satisfaction

Traditional service metrics remain useful, but they do not explain the full customer experience. Leadership should also examine what customers actually do: whether they contact the company repeatedly, abandon journeys, bypass designed processes, or lose confidence in the experience. Telemetry and AI can reveal where effort, confusion, and friction occur. The purpose is not to produce another dashboard. It is to understand why a problem is happening and what the organization should change. Metrics become strategically useful only when operational evidence is connected to customer and commercial outcomes.

Use AI to elevate service, not simply to reduce cost

AI can make customer service faster, more proactive, and more valuable to the business. The strongest results will not come from automating the largest number of steps. They will come from choosing the right use cases, preserving human involvement where it matters, and connecting service data to business outcomes. Companies that align data, accountability, and technology can move customer service beyond ticket resolution. It becomes a management capability that builds trust, protects revenue, and strengthens the full customer relationship.

Source: Elevating Customer Service to a Strategic Imperative with AI, Harvard Business Review Analytic Services, 2026.