~/writing/architecture-meets-sow scoping 5 min read
Jonatan Reiners Solutions Architect :: Premier AWS Partner
~/writing / architecture-meets-sow
#scoping 2026 — 04 — 02 5 min read

Where the architecture meets the SoW

Effort estimation when technical reality collides with a fixed price — and how I stop the gap from becoming my problem.

Engineers don’t scope and salespeople can’t architect, so the statement of work gets written in the seam between them — and the seam is where fixed-price delivery quietly bleeds. The architecture assumes things the SoW never priced; the SoW promises things the architecture can’t hold. When they diverge mid-delivery, the gap defaults to the delivery team. Here’s how I stop that.

The move: assumptions are line items, not prose

Every assumption that changes the price gets written as a named, testable clause tied to a number — not buried in a paragraph of throat-clearing. “Source database is < 2TB and on a supported version” is testable on day one. “Customer provides timely access” is not; it’s a future argument.

An assumption you can’t test on day one is a dispute you’ve scheduled for month three.

Three clauses that earn their place

  • The scope boundary. What’s explicitly out. The out-list prevents more scope creep than the in-list ever will.
  • The trigger. The condition that converts an assumption into a change request — stated, with the rate, before anyone’s emotional about it.
  • The count cap. “Up to 40 applications” turns an open-ended estate into a priced one. App forty-one is a conversation, not a crisis.

Why this is the architect’s job

Because only the person who understands the architecture knows which assumptions are load-bearing. The commercial team can’t see that the innocuous-looking integration is the one that sinks the timeline. Scoping the seam is the most commercial thing a technical SA does — and the cheapest insurance the delivery team will ever buy. The format I hand over is in the next post.

$ filed under scoping — next: Estimate in dollars → →