The first full cycle of reporting under the European Sustainability Reporting Standards is behind us, and the honest summary is that almost everyone underestimated it. Not the writing, the plumbing. Teams that had produced voluntary sustainability reports for years still found themselves rebuilding half their process from scratch, because ESRS asks for something different: structured, defensible, assurance-ready data rather than a well-designed narrative.

We supported several disclosures through that first cycle, a number of them in cement and other hard-to-abate sectors. A few lessons showed up again and again, regardless of company size or how mature the sustainability function looked on paper.

Double materiality took longer than anyone budgeted

The double materiality assessment is the spine of an ESRS report, and it is where the most time quietly disappeared. Teams treated it as a workshop to be run once, when in practice it is an evidence trail that has to survive an assurance provider reading it backwards. Impacts, risks and opportunities need to be tied to real thresholds and real stakeholder input, not assigned in a spreadsheet the afternoon before sign-off.

The teams that struggled least were the ones that started the assessment early, wrote down their methodology as they went, and could show why a topic landed inside or outside the boundary.

Data collection is a plumbing problem

The second recurring theme: most of the effort was not in reporting the numbers, it was in being able to trust them. Datapoints that had never left an operations team now needed owners, definitions, and a chain of custody. Value-chain data, especially Scope 3 and workforce figures from contractors, was the hardest to pin down.

The report is the easy part. The year is won or lost on whether the data underneath it can be traced.

The gaps assurers actually flagged

Across the disclosures we saw reviewed, the same handful of issues came back from limited assurance again and again:

None of these are exotic. They are the predictable result of a first cycle run at speed, and every one of them is fixable before the next.

If you do one thing before next cycle

Build a datapoint register: every required disclosure, its owner, its source system, and its calculation method, in one place. It is unglamorous, and it removes most of the year-two panic.

What changes for cycle two

The good news is that the heaviest lifting, the materiality methodology and the data architecture, is largely one-time work. Cycle two is about maintenance and depth rather than reinvention: refreshing the assessment, tightening estimates into measured figures, and widening value-chain coverage.

Teams that documented their first cycle well will feel the difference immediately. Teams that treated it as a one-off deliverable will, unfortunately, repeat much of the scramble.

The takeaway

ESRS rewards process over polish. The companies that came through the first cycle in good shape were rarely the ones with the glossiest reports; they were the ones who could answer, without hesitation, where every number came from and why every topic was in scope. That is the standard the second cycle will be measured against, and it is worth building toward now.