The end of cost-to-serve competition

By Marc Sauthoff and Lennart Lohrisch

Cost-to-serve in energy retail is no longer falling incrementally but structurally: agentic AI, smart meter data, and cloud-native platforms are pushing the marginal cost of a customer transaction toward zero. But when all providers converge on a similar cost level within a few years, process efficiency loses its differentiating power. Competition shifts to data quality, customer experience, and speed of scaling. Consequently, the attractiveness map is being redrawn. Segments considered uneconomical today – small business, prosumers, dynamic tariffs, switch-prone customer groups – become profitable. Those who occupy them first grow into spaces competitors still avoid through a legacy cost lens.

The end of a 25-year paradigm

Since market liberalization, energy retail has followed a stable logic: margins per customer are thin, so cost-to-serve determines profitability. Whoever could serve a household customer for EUR 25 a year instead of EUR 50 could work segments that were structurally unattractive for others – or survive the price war. The discounters perfected this game, and the incumbents spent two decades catching up: shared services, nearshoring, IVR systems, customer portals, robotic process automation.

Each of these waves delivered 10-20% in cost reduction – and each was eventually exhausted, because in the end a human always had to touch the case. Digitalization shifted contacts but rarely resolved them. The agent remained the limiting resource, and customer growth meant linear capacity build-up. This assumption is now collapsing. AI changes the economics because customer service is no longer fundamentally constrained by human labor.

The result: cost-to-serve is now falling structurally, because the marginal cost of a customer contact or an exception case, for instance, is heading toward zero. This is a regime change: the technology behind it is available to every provider as standard software and AI tooling. But when all competitors converge on a similar cost level within a few years, process efficiency loses its differentiating power. It becomes the entry ticket – just as no one advertises with a customer portal anymore.

The strategic point lies in what comes next. Competition shifts to three new fields: data quality, because it determines the automation rate. Customer experience, because it is the last perceptible difference once price and process converge. And speed of scaling, because growth is possible for the first time without proportional cost build-up.

And there is a second, often overlooked consequence: segments considered unattractive today become profitable. The overview below shows how the assessment of key customer groups shifts once service costs no longer grow linearly with the customer base.

Structural forces reshaping energy retail

Four drivers are converging to fundamentally change the economics of energy retail. Each is significant on its own, but together they create a step change in how customer service, operations, and technology can be organized. What was previously constrained by labor-intensive processes becomes increasingly software-driven. Four drivers are working in concert, from agentic AI to SaaS solutions, while a fifth factor explains why data quality, of all things, becomes the new bottleneck.

1. Agentic AI resolves cases instead of rerouting them

Digitalization to date has shifted contacts – from phone to portal, from letter to app. It has rarely resolved them. AI agents based on large language models now handle cases end-to-end for the first time: payment plan adjustments, relocations, invoice clarifications, tariff advisory – around the clock, multilingual, at marginal costs near zero. The relevant steering metric becomes the straight-through processing rate across the entire case, no longer handle time in the contact center.

2. Smart meters and market communication create the data foundation

The rollout of smart metering systems, the reduction of supplier switching to 24 hours, and the progressive standardization of market communication make processes machine-readable that previously lived on exceptions and manual clarification. What used to be the graveyard of exception cases in retail operations becomes automatable – provided the master data is correct.

3. Cloud-native platforms commoditize the billing stack

SaaS solutions make first-class billing and service processes available as a standard product. The scale advantage of large IT organizations erodes; fixed costs become variable. A municipal utility with 80,000 customers can operate at the same process level as a large group – process excellence itself is no longer a moat.

4. Software scales, headcount does not

Until now, customer growth meant linear build-up of service capacity – the real reason low-margin segments were unattractive. In an AI-first operating logic, customer base and service costs decouple. Break-even per customer falls so far that the attractiveness map of segments has to be redrawn.

The flip side: Data quality becomes the limiting factor

Automation exposes data quality deficiencies mercilessly. An experienced service agent compensates for flawed master data with contextual knowledge – an AI agent turns it into an exception case or a wrong answer. Data quality thus moves from hygiene topic to the core of the cost position: it directly determines what share of cases can be automated end-to-end.

Strategic priorities for utilities

The new economics of utility retail require a different strategic playbook. Some actions are becoming fundamental to remain competitive, while others create opportunities to build lasting advantage. The priorities fall into two agendas: securing the capabilities every utility will need, and investing in the differentiators that will define tomorrow's market leaders.

Staying competitive: Securing the entry ticket

• Redesign the highest-cost processes AI-first.
• Build the data foundation automation depends on.
• Modernize the billing and service stack AI-scale operations.

Becoming a leader: Build tomorrow’s competitive advantage

• Design for zero-inbound, not efficient inbound.
• Expand into customer segments made profitable by lower service costs.
• Reinvest productivity gains to accelerate growth and improve customer experience.

The coming years will be defined by who redesigns their operating model around AI. As process efficiency becomes a prerequisite rather than a differentiator, competitive advantage shifts to data quality, customer experience, and the ability to scale faster than competitors. Utilities that view these changes as incremental digitalization risk competing on yesterday's economics. Those that recognize the structural shift early can reshape both their cost position and their growth trajectory.

Sign up for our newsletter and get regular insights on newest publications related to Energy & Utilities.