Hearth

The Warm Homes Plan commits £15bn to upgrade up to 5 million homes and lift up to 1 million families out of fuel poverty by 2030. The private rented sector must reach EPC band C by October 2030. Area-based delivery begins in 2027/28, tied to local energy plans.

There is no unified eligibility framework behind it. Households are reached through separate scheme-specific routes, and ECO Flex still depends on individual councils issuing their own declarations — producing roughly three hundred inconsistent eligibility regimes for a single national objective.

The targeting data is weaker than it looks. Energy Performance Certificates only cover properties transacted since 2008, so a large share of the housing stock has no certificate at all; scores are modelled rather than measured; and certificates last ten years, so many are stale. Nobody can currently target the un-certificated half of the stock.

The absence of outcome verification has already produced a scandal. Around 65,000 households received solid wall insulation under recent schemes; 39 businesses were suspended and Ofgem was ordered to check every installation and write to all affected households. Defects included missing ventilation and exposed insulation. That failure surfaced through complaints, not data — because nothing checks whether an installed measure produced the predicted saving. Government has since said the scheme responsible was "the source of the majority of poor-quality installations."

The plan has a budget and no data plan. Searching the published Warm Homes Plan for data sharing, data matching, DWP, HMRC and Digital Economy Act returns zero hits in every case. Its only targeting commitment is hedged as a "potential opportunity", with no legal gateway named, no timeline and no funding line.

Worse, the legal plumbing points at the wrong actor. The data-sharing gateway that makes household-level targeting lawful was built to push government data to energy suppliers, so they could deliver supplier obligations. Those obligations are being abolished — one scheme ended in March 2026, the other ends 31 December 2026, with no successor. Delivery moves to local authorities and a new agency, and local authorities have no equivalent route to benefits data. The current eligibility checker is a self-service funnel relying on self-declared income, which government concedes "will sometimes tell users they are ineligible when they are eligible."

And the evidence base is fracturing beneath all of it. Certificates lost their primary key in March 2026 — government states plainly that "it is not possible to link certificate numbers to the former LMK_KEY" — the headline rating is replaced by four separate metrics from the second half of 2027, and the underlying calculation engine is being replaced too. Three discontinuities in a single programme.

Meanwhile 2.7 million households in England are in fuel poverty, the aggregate gap is £1.11bn, and cold and damp housing costs the NHS an estimated £900m a year.

Hearth is deliberately not a targeting tool. Targeting requires household income and benefits data that cannot lawfully be published, so it belongs to councils and scheme administrators who hold it — and commercial products already serve them, at £20,000 to £80,000 a year. Building a public rival to those would duplicate a served market and run into a legal wall.

What nobody does is hold the programme to account for its outcomes. Hearth tracks post-installation certificate re-lodgement against the uplift each measure predicted, by measure type, by scheme and by installer. Underperformance surfaces as a signal within months rather than as a scandal after 65,000 installations.

It publishes the spend-to-outcome record: which areas received investment under which scheme, at what cost per property, and what measured improvement followed. That is the record a public accounts committee needs and cannot currently obtain, and it exists nowhere for a £15bn programme.

It also fills a genuine data gap that helps everyone including the commercial tools — imputing energy performance for the substantial share of stock with no certificate, published openly with explicit uncertainty. Properties that have not been sold since 2008 are currently invisible to every scheme, which means the programme systematically misses homes for reasons that have nothing to do with need.

Throughout, modelled and measured values are flagged distinctly. A system that hides the weakness of its own inputs is how £15bn gets spent on the wrong houses and reported as a success.