Eleven years in environmental statistics before Noverde, where she settles the boundary question that decides what every later figure means.
The boundary question is the first one we ask and the one most engagements would rather skip. Which sites are yours to report. Which joint ventures. Which leased assets sit on your inventory and which sit on somebody else's. Under the GHG Protocol more than one answer is usually defensible, and that is precisely why the answer has to be chosen deliberately rather than arrived at by accident.
A boundary agreed halfway through a measurement invalidates everything counted up to that point. So it is settled in writing, with the people who will have to defend it, before a single meter export is opened.
Four pages, usually, and they are the four pages an assurer reads first. They are also the ones a client is most tempted to leave until the end, which is exactly backwards: every figure counted afterwards depends on what these four pages happen to say.
A verbal boundary drifts. Twelve months later nobody remembers whether the Rotterdam site was in or out, and the prior year is no longer a comparative. A signature with a date on it costs an afternoon and saves a restatement.
Half the restatements I have seen were not measurement errors at all. They were a boundary that moved while nobody was looking at it.
Materiality. A topic that scores below the line still gets written down, with the reason it was set aside, and that list is published. Clients find this uncomfortable, because it hands a reader the exact place to push. That is what makes the rest of the matrix worth reading.
