A weak Annex VI audit usually fails before the recommendation stage because the data boundary is not traceable. The audit file should show where each consumption value came from, what period it covers, which meters, invoices, operational systems, or load-profile exports support it, and why the selected sites, operations, installations, or transport activity are representative.
Representativeness matters because Annex VI requires enough coverage to draw a reliable picture of overall energy performance and to identify the most significant improvement opportunities. Annex VI does not set one EU-wide coverage percentage or sampling formula. Commission Recommendation (EU) 2024/2002 describes possible Member State de minimis, clustering, and sampling approaches, including similar-site samples, but those examples are non-binding guidance. The applicable national scheme decides whether a minimum percentage, sample size, or specific method applies.
Where sampling is allowed, the sample needs a reason: energy intensity, operating hours, process type, building use, climate or occupancy pattern, production line, fleet type, or another factor tied to the energy profile. Keep the cluster characteristics, omitted population, selected sample, and representativeness rationale in the report.