Reference models in the Preliminary Phase are usually presented as neutral infrastructure -- a taxonomy of technology components, a map of integration patterns, shared vocabulary agreed before the real work starts. In a commercial context that framing holds, because the endpoint metric is already well-defined: revenue, margin, growth. You build the reference model to serve that measurement.
The question I keep returning to is what the reference model looks like when the endpoint metric is community access. Not API calls per second or uptime percentage, but whether a kura now has something it did not have before. The standard TOGAF Technical Reference Model is structured around platform services and application software. There is no row for schools reached or organisations enabled. That gap is worth naming early, because it shapes what Phase C information systems work has to carry. If the reference model does not include mission-delivery data structures from the start, we will retrofit them under pressure later -- and retrofitted metrics are usually wrong.
The Preliminary Phase decision is not just which frameworks we adopt. It is whether those frameworks can carry the right signals. The architecture capability and framework docs (full docs: https://cloud.nzrtnetwork.com/nextcloud/index.php/s/BrfYrN4PxXgPW5t | https://cloud.nzrtnetwork.com/nextcloud/index.php/s/dg8n9mRTt9FTQrW) set the foundation for that. Before we move into Phase A, I want one thing confirmed: does our chosen reference model have a designated place for mission metrics, or does it need to be extended? The answer changes what the information architecture looks like at every subsequent phase.
