The customer who never calls again

By Lennart Lohrisch and Steffen Thiel

AI assistants are becoming the new decision-makers in energy retail. Instead of comparing tariffs, switching suppliers, or resolving complaints themselves, customers will increasingly delegate these tasks to software acting on their behalf. This creates a new business-to-agent (B2A) channel in which utilities no longer compete for human attention but for machine-readable visibility. Success will depend on structured product data, API-enabled processes, and seamless machine-to-machine interactions rather than marketing campaigns or call-center excellence. As switching friction disappears, customer retention can no longer rely on convenience. Utilities must become accessible to AI agents, or risk disappearing from the customer's decision process altogether.

The customer who never calls again

For two decades, energy retail built its channels for humans. First the call center, then the web portal, then the app – every expansion stage was designed around a person who reads, clicks, waits, and can be persuaded. The entire sales and service toolkit is calibrated to this assumption – campaigns, landing pages, win-back calls, bonus banners.

This assumption is becoming obsolete. Customers are beginning to delegate their energy affairs to personal AI assistants: check my electricity tariff, switch if it pays off, dispute the invoice if something is wrong. The assistant does not compare like a human – it reads no advertising banners, feels no switching inertia, forgets no cancellation deadline, and is not impressed by a beautifully designed landing page. It queries terms machine-to-machine, evaluates them against its principal’s preferences, and closes the deal – or not.

This creates a fundamentally new channel in energy retail: business-to-agent. The utility no longer negotiates with the customer but with their digital representative. And this channel has an uncomfortable property: it is binary. Whoever is machine-readable – with structured tariff data, queryable terms, API-enabled processes for switching and complaints – exists for the agents. Whoever has just a call center and a web portal is invisible. An offer the agent cannot process simply does not appear in its selection.

The historical parallel makes the urgency tangible. When price comparison platforms emerged, many utilities initially treated them as a fringe phenomenon, until they became the dominant acquisition channel and dictated terms. B2A (business-to-agent) is the next escalation stage of the same development, only more radical: the platform helped the human compare; the agent replaces the human across the entire transaction. And unlike the platform era, there is no intermediary from whom visibility can be bought – the interface must be provided by the utility. Those who learned the platform lesson too late lost margin. But those who learn the B2A lesson too late lose access to demand altogether.

The comparison below shows how fundamentally the sales logic shifts when it is no longer the human but their agent who decides.

The forces creating the B2A channel

The emergence of B2A interactions results from several developments reinforcing one another: increasingly capable AI assistants, consumer willingness to delegate routine decisions, and a regulatory environment that makes energy markets more digital, standardized, and machine-readable. Together, these forces are creating the conditions for a fundamentally new customer channel.

Consumer assistants are reaching operational capability

The major technology providers are equipping their assistants with agentic capabilities: they no longer merely research, they execute – booking, canceling, comparing, disputing. With standardized agent protocols and agent-capable browsers, the technical infrastructure is emerging through which an assistant operates on third-party services on the customer’s behalf. Energy contracts are an ideal use case: standardizable, data-driven, emotionally uncharged, with clear savings potential.

The energy contract is the perfect delegation product

Customers experience tariff comparison and switching as a tedious chore with no experiential value. Precisely such low-involvement tasks are delegated first. The switching inertia that underpinned incumbent margins for decades was never loyalty – it was convenience. The agent eliminates the convenience barrier entirely: switching will cost the customer zero effort, because they no longer do it themselves.

Regulation accelerates machine-readability

The required fast supplier switching makes the switching process fully digital and quasi-instantaneous. The standardization of market communication, smart meter data, and the European data economy agenda – the Data Act, the right to data portability – create the foundations on which third-party agents can legitimately query customer data and contract terms. Regulation is unintentionally building the infrastructure of the B2A channel.

The trust relationship inverts

Customers increasingly trust their assistant more than the provider – the assistant pursues no sales interests, the utility does. The loyalty relationship shifts accordingly: it is no longer the utility that owns the customer relationship, but the agent. From the customer’s perspective, the utility becomes an interchangeable supplier behind an interface controlled by someone else.

Moving from visibility to leadership

As AI assistants become a new customer interface, utilities must prepare for two distinct challenges: becoming accessible to agents and competing successfully within the new channel. The first is about securing visibility; the second is about shaping the future of customer relationships.

Staying competitive: Becoming visible to agents

• Become machine-readable and transactable. Publish structured, complete, up-to-date product and tariff data, and enable end-to-end API journeys for quote retrieval, contract switching, and complaints. Machine readability becomes the digital storefront of the B2A era.
• Measure and understand agent traffic. Measure the share of interactions generated by AI agents and monitor its growth. Understanding agent traffic will become as important as tracking digital adoption today.
• Clarify the governance of the agent mandate. Define authentication, mandates, and liability frameworks for AI agents acting on behalf of customers.

Becoming a leader: Shaping the channel

• Build a genuine agent experience. Go beyond API availability by providing fast, reliable, consistent interfaces that AI agents prefer as they simplify decision-making.
• Develop agent-native products. Create machine-negotiable tariffs, dynamic pricing models, and product bundles that AI agents can evaluate and optimize.
• Own the customer interface. Offer a utility-operated energy assistant that manages tariffs, PV, storage, and EV charging before third-party agents become the primary customer interface.

AI assistants are already part of the customer journey – the question now is how quickly they will become the primary interface for routine energy decisions. As B2A interactions gain traction, visibility will depend less on marketing and more on machine-readable products, processes, and data. Utilities that adapt early can shape this new channel. Those that do not risk becoming invisible – not because they offer the wrong product, but because the agent never considers it.

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