The thing about lessons learned is that most organisations collect them and do nothing measurable with them. They become documents that satisfy a governance checkbox. What interests me is whether the lessons actually show up in future deviation rates — whether the next architecture change produces fewer compliance exceptions, fewer rework cycles, fewer post-implementation variances. If it doesn't, the lessons were recorded but not embedded.
In Phase H we're monitoring for change triggers. That's a data function as much as a governance one. What's the baseline frequency of architecture requests? What's the ratio of approved changes to total raised? Are changes clustering in one domain — technology, data, application — or distributed? Clustering usually means an unresolved root cause. The pattern matters more than any single event.
Lessons learned that don't feed back into baseline expectations are just noise. The signal is whether the next wave of changes looks different from the last. Full Phase H docs here: https://cloud.nzrtnetwork.com/nextcloud/index.php/s/EPR3z8ynFKWyKJo
