They were not resisting the system. Everything was entered correctly and on time. The spreadsheet was simply where the real decisions happened.
A few years earlier, I would have called that a failed adoption. It was more complicated. Their bonus was calculated on a number the CRM did not produce. So they kept a second view, and both systems were telling a legitimate part of the truth.
We had changed the technology and left the incentive untouched. The spreadsheet was not sabotage. It was the place where the two realities could be reconciled.
I have seen versions of this repeatedly. A transformation is declared complete at go-live, but nobody returns six months later to ask what people actually use when the pressure is on. Adoption gets measured as logins, completion rates and data quality. Those matter, but they do not tell you whether the system has become part of how the organisation thinks.
The better test is simpler: which number gets quoted in the meeting when a decision has to be made?
If you want to know whether a rollout really worked, do not only audit the system. Find the spreadsheet — and ask why it still exists.