Claim substantiation register

Evidence & Methodology

We make ambitious claims about building performance. This register is how we hold ourselves to them. Take any figure from our marketing, follow its citation here, and you will find the method that produced it and the limits of what it proves.

  • Every headline claim we publish is listed here with its methodology, sample size, baseline, counterfactual, measurement boundary, uncertainty range, engine version and validation result.
  • Where the evidence supports only a narrower version of a claim, the card states the permitted wording and the wording we do not use.
  • Where a claim is not yet substantiated, it is marked as such and withdrawn from our marketing until it is.
Independent regression register
Validation & regression runs

How this register relates to our Terms

Our Terms of Service require that platform outputs are verified and signed off by a suitably qualified professional before they are relied upon. That requirement stands, and this register does not displace it. What the register does is remove the ambiguity: rather than describing outputs only as indicative, we publish the method, the sample and the measurement boundary behind each figure, so a reader can judge for themselves what the number proves and where professional judgement is still required.

Where a claim on this page is marked under review or withdrawn, it is removed from our marketing until the evidence supports it. If you believe a claim we publish is not substantiated by the evidence shown here, tell us and we will correct the claim or the evidence.