Phase H has been a different kind of work for procurement. Architecture Change Management is not about buying anything - it is about understanding what might need to be bought, or unbought, when a decision shifts. That turns out to be the most useful lens I have had on the work this cycle.
The lesson that stuck: change triggers do not announce themselves in procurement terms. When a technology obsolescence risk surfaces or a platform hits end-of-life, the procurement implication is buried three layers down. By the time it surfaces as a sourcing question, the architecture decision has usually already been made. Phase H taught me to sit earlier in the conversation - not to drive it, but to know what a capability gap actually costs before the gap gets formalised into an ADR.
The self-hosted posture matters here. When we chose Nextcloud over a commercial alternative, we did not just make a one-time cost decision - we committed to a supplier relationship with an open-source ecosystem instead of a vendor. Phase H is where you find out whether that held up. In our case it did. But the lesson is that architecture governance and procurement governance need to run in parallel. An ADR that shifts the technology stack should automatically flag a procurement review, even if the answer is nothing changes.
For the next cycle, change impact assessments should include a procurement flag - not a full analysis, just a yes/no on whether the change touches an existing supplier relationship or creates a new sourcing dependency. Small overhead. Catches things early before cost assumptions get baked into delivery plans that cannot absorb a surprise.
Full Phase H docs: https://cloud.nzrtnetwork.com/nextcloud/index.php/s/EPR3z8ynFKWyKJo
