The data revolution the UK
energy debt crisis demands

Intro

The numbers are stark. Ofgem reports that UK domestic energy debt has reached a record £4.5 billion, owed by approximately three million households — a 250% increase in five years. Latest sources put this figure closer to £5.6bn and industry forecasts suggest this could climb to £7 billion by the end of 2026, adding an estimated £10–£15 per year to every customer's bill as costs are mutualised.

While a host of new proposals, including Prime Minister Andy Burnham’s pledge to remove value-added tax from domestic electricity bills, possible introduction of new fees on data center projects and a potential decoupling of electricity and gas prices may help to limit future consumer costs, but they do not address the underlying challenges utilities face with unpaid bills. This is no longer a collections problem. It is a systemic failure that redistributes cost onto those who do pay, while leaving those who genuinely cannot pay increasingly exposed.

The scale of the numbers, sobering as they are, can obscure a more important truth: the debt population is not homogeneous. Approximately 30% of customers are "in debt" at any given moment, but the reasons why vary enormously. Some cannot pay. Some will not. Some are caught in a grey zone shaped by circumstance, behavioural factors and inadequate support. Some have fallen behind because regulation designed to protect the vulnerable has inadvertently created cover for those with the means to pay.

As EDF's recent analysis illustrates, energy debt owed by customers with high affordability scores has risen by almost 10% in a single year, while debt among those with low affordability scores has remained broadly flat. This is not a failure of customer intent; it is a failure of the system to sufficiently distinguish between need and non-payment.

Diffusing this volatile situation is the central challenge utilities face today.

Moving toward a household view of financial health

To address it, utilities need to change the way they think about their customer bases, and, increasingly, focus not just on the single person whose name is on the bill, but on the entire household and the multitude of variables that come along with it. A single-account, single-person view of debt is fundamentally limited. It misses the dynamics of shared households. It fails to capture the full picture of a home’s historical energy record and does not address many underlying symptoms of financial stress. Importantly, it cannot distinguish between a genuinely struggling family and a household where one member is gaming the system while others have capacity to pay.

The solution is to shift the utility’s understanding of energy debt away from the individual and toward the property as the fundamental unit of analysis. By creating an "open book" at the property level – a persistent, portable record that travels with the address rather than the account holder – utilities would be able to surface patterns that are currently invisible: repeated arrears across successive tenants, properties with smart meter issues, chronic affordability challenges regardless of occupant, or households where vulnerability indicators are consistent and should trigger proactive support across a number of organisations.

This is not a small idea. It would require a significant shift in how data is collected, stored, and shared across suppliers, and a frank conversation about what is permissible and what is genuinely useful. But the potential upside is transformative. Suppliers entering a new customer relationship would do so with relevant context. Support could be offered before crisis hits. And wasted cost – the repeated cycles of assessment, escalation, and write-off that currently burden supplier cost-to-serve – could be dramatically reduced.

Data needs to do the heavy lifting

To do this effectively, utilities would need to be able to leverage external and internal data for better decision-making. Today, most industry affordability assessments built on a foundation of unreliable data that often lead to misclassification and lack of visibility into the real underlying financial health of a customer base – in most cases resulting in an industry average accuracy rate of just about 10%. That simply will not cut it.

A richer, more joined-up data ecosystem combining internal account data, third-party sources, and crucially government-held data is needed to enable a level of precision that the current framework simply cannot achieve. Government departments hold information about benefits, income, and vulnerability that, if accessible to suppliers in an appropriate and consented framework, could transform the accuracy of affordability assessments. This is not without complexity – policy, privacy, and trust considerations are all very real issues, but the principle is sound and the precedent exists in other sectors.

Beyond government data, the use of AI to interrogate household-level datasets more effectively by identifying behavioural signals, clustering households by risk profile, and predicting which accounts are likely to deteriorate before they do represents a significant opportunity. Predictive analytics, deployed with the right data inputs, can shift the industry from reactive debt management to proactive support.

Put simply, utilities need more data and more robust analytics capabilities to develop more tailored, personalized outreach strategies that connect with customers. Personalisation strategies that account for a customer's full circumstances including their household composition, their history, their communication preferences, their demonstrated engagement, can transform the nature of the supplier-customer relationship in arrears. Customers who feel seen and understood are more likely to engage. Those who feel processed are more likely to disengage. The evolution of societal attitudes to debt, flagged in the roundtable, makes this more important, not less: a generation that views debt differently requires an industry that responds differently.

The window of opportunity is now

Winter will intensify every pressure described here. The price cap rising, household budgets are already stretched, and the debt overhang is at a historic high. London, in particular, is a hotspot requiring targeted, place-based intervention.

The opportunity and urgency is to act now. That means accelerating work on data-sharing frameworks, engaging government on data access and a genuinely open vulnerability register, developing the evidence framework for customer verification, and piloting AI-driven segmentation approaches that can distinguish need from non-payment with genuine accuracy.

The Debt Relief Scheme, long delayed and repeatedly diluted, also needs to be delivered at meaningful scale. Legacy debt is a drag on the entire system. Clearing it, for those who genuinely accumulated it through no fault of their own, creates a set of conditions for a reset that benefits everyone.