When a Rating Does Not Move

A rating that does not move after work is disappointing, but it is not always proof that work failed. Ratings move only when modelled inputs change enough to move the score across a band threshold, or at all, and several factors can prevent that: work that improves comfort but not modelled inputs, work recorded as a default rather than evidenced, a score improvement too small to cross a band, or a method change offsetting a physical improvement. Each cause needs a different response.

The first step is therefore to compare inputs, not bands. If old and new certificates show the same wall, heating or glazing inputs despite work, the assessment missed the work. If inputs changed but the score barely moved, the model values that measure modestly for this home. If the score moved but the band did not, the work helped but not enough to cross a threshold. Those three situations look identical as bands, but they mean very different things.

Band thresholds are coarse. A score improvement of several points may still leave a home in the same band if it started far from a boundary, while a one-point improvement may move a home across a boundary if it started just below it. Band stability therefore does not prove work worthlessness, and band movement does not prove work value; score movement is the finer measure, and it should always be checked alongside the band.

Model limits also matter. Some measures improve comfort, air quality or resilience without much modelled energy effect: draught-proofing a single door, improving controls usability, or fixing a cold bridge in one room may transform lived experience while barely moving a whole-home model. Those measures are not failures because the model values them modestly; the model is a whole-home average, and lived comfort is room-specific.

Compare input summaries first. List wall, roof, heating, glazing and control inputs on old and new certificates side by side, and mark which changed and which did not. If a completed measure does not appear as a changed input, gather evidence, photographs, invoices and product data, and ask the assessor to revisit that input, because evidence is the usual missing link.

Check score movement, not just band movement. If the score improved, even within the same band, work had a modelled effect, and further measures may move the band in time. If the score did not move at all despite input changes, ask the assessor to explain the model's treatment of that measure for this home, and whether a different measure would have a larger modelled effect.

In the UK, EPC method updates can offset physical improvements: a home improved between certificates issued under different method versions may show little rating change because the newer method is stricter or uses different assumptions. A UK reader comparing certificates across method versions should check method dates and ask an assessor whether method change, rather than work failure, explains a static rating.

In EU member states, method and factor changes have similar effects, with primary energy factors and class boundaries revised nationally over time. A home improved in a country that tightened its classes may show no class movement despite real improvement, and that outcome should be judged on physical and delivered energy evidence, not on class alone.

One mistake is repeating work because a rating did not move. A household that insulates further, or replaces a system again, in response to a static band, without checking inputs and score, may spend on a measure the model already values, while the real cause, such as missing evidence or a control fault, remains unaddressed. Diagnose the static rating before spending more.

Another mistake is dismissing work that improved comfort but not rating. A draught-proofed, warmer room with steadier temperatures is a real improvement, even if a whole-home model barely registers it, and dismissing it because a letter did not change undervalues lived experience. Judge work on comfort, meter data and rating together, not on rating alone.

Keep old and new certificates, input comparisons, score movements, evidence packs, and comfort and meter notes before and after work. That combined record allows any future professional to see whether a static rating reflects work failure, model limits, evidence gaps, or method change, and to advise accordingly.

Ask the assessor to explain, input by input, why the rating did or did not move, and what evidence or further measure would move it. Ask also whether a different assessment, such as a detailed audit, would capture benefits the rating model misses. Those answers turn a disappointing letter into a diagnostic opportunity.