McKinsey described telecommunications as a sector at an inflection point. In “Telcos’ AI inflection point: What leaders do to capture value,” the firm argues that agentic AI could help operators reverse the weak trajectory of recent years, but only if they avoid adding yet another isolated technology layer. The core claim is straightforward: value will not come from the model alone, but from an organizational redesign that connects network operations, customer care, commercial teams, and finance.
McKinsey points to Japan’s NTT DOCOMO as a concrete example. Working with advisers, the company developed the Customer Network Experience index, an AI-based metric that translates network quality into something different parts of the business can actually use. That matters because traditional telecom data may be technically accurate while remaining commercially weak. Once network performance is translated into churn risk, revenue per customer, and service quality, AI stops being an experiment and becomes a management tool.
The article also addresses a broader structural problem for operators: heavy infrastructure spending, price pressure, and weak growth. In that environment, it is not enough to promise savings in call centers or faster internal analysis. Companies must show that AI improves retention, reduces network intervention errors, speeds up fault resolution, and sharpens capacity decisions. Otherwise even a large investment will dissolve into operating cost without a visible effect on revenue.
For executives, the important point is that McKinsey does not treat AI as a standalone technology strategy. Instead, it shows that operators need new ownership rules for data, model decisions, and business outcomes. “The network is the product” is not a slogan here; it is an accounting fact. If customers experience outages, slowdowns, or poor complaint handling, that hits margin, churn, and acquisition cost directly.
The article matters beyond telecom as well. In capital-intensive sectors, AI return will depend largely on whether companies can connect operational data to commercial decision-making. Businesses that keep AI inside technical teams will get better models. Businesses that push it into revenue management will get a better business.