The thing about a wrong architectural decision is that it is rarely felt in year one. In year one, everything still moves. The product ships. The model trains. The dashboard renders. The decision is buried under the surfaces that followed it. The building is up. Nobody is questioning the section.
Year two is when the friction begins. New surfaces require renovations they were not supposed to require. Decisions that should have composed get rebuilt instead. The team starts calling them 'edge cases' — a phrase that almost always means the architecture did not anticipate a load it should have.
Year three is when the cost becomes visible at the bottom of the org chart. Velocity falls. Engineering leadership stops promising calendar dates. The phrase 'we should have done this differently from the start' enters the meeting. The team is now load-bearing the wrong decision — and pulling it out would mean rebuilding the strata above it.
Year three is what the building looks like when nobody sat with the section.
What the studio sees from the field is that the wrong architecture is almost never identified as wrong at the time it is made. It is identified as 'good enough for now' or 'we can change this later.' Both phrases are tells. Both are the engineering equivalent of building a foundation on assumptions the building is not yet aware of.
The right time to draw the section is before the building goes up. The second-best time is at the moment of an honest re-section — pause, draw what is actually load-bearing, name which decisions can be replaced and which cannot. The team that does this in year two saves the decade. The team that waits for year three to be obvious does not.
The studio's working definition of architectural maturity is the willingness to sit with the section for as long as it takes to be honest. Not as long as it takes to feel productive. Not as long as it takes to look senior. As long as it takes to be honest. Year three is what the building looks like when that sitting was skipped.