The first thing I do on a new site is look for the reason someone else walked away from it. There is almost always one, and it is almost never the price. A lot too shallow to hold a rentable floor plate. A frontage that forces the core against a wall it cannot sit against. An existing structure that cannot be removed and cannot easily be built over.

I take those sites on because the reasoning around them is where the real work is. A site with no constraints produces a building that any competent developer could have produced, and the result tends to show it.

What I actually do on a project is hold the order of decisions. Hard buildings are not solved all at once; they are solved in sequence, and the sequence is usually the whole insight. Something has to be true before the next thing becomes possible, and getting that order wrong is more expensive than getting any single decision wrong. A structural strategy chosen before the review path is understood can be legal and unbuildable at the same time.

I think the most common mistake in this business is treating design, engineering, cost and approvals as four conversations that happen in turn. They are one conversation. A facade decision is a structural decision and a schedule decision, and pretending otherwise just moves the cost somewhere it will not be noticed until later.

That is a position a competent developer could argue with, and some do. The counterargument is that separating those functions produces sharper specialists and cleaner accountability, and on an ordinary building that is probably right. My experience is that it stops being right as soon as the constraints get unusual, because the tradeoffs then run across the boundaries rather than inside them.

The honest limit is that this way of working is slow at the front end and it looks like indecision to anyone watching from outside. Months go into questions that produce no drawing anybody can point at. On a project with an obvious answer that time is wasted. On a project without one it is the only part that matters, and I have been wrong about which kind of project I was standing on.

Nearly every problem set out on this site was actually solved by somebody else: a structural engineer, a facade fabricator, a preservation consultant, the superintendent running the job. I name that because a developer who quietly absorbs that work into his own account of things gradually stops being told the truth by the people doing it, and that bill arrives years later.

Most of what is written here comes out of the work at JDS Development Group. The project that taught me the most about sequence was the one whose footprint was settled by a courtyard before a single other decision could be taken.

The reason I bother writing any of this down is that the reasoning is the part that travels. A building belongs to its own lot and nobody else is ever going to build it. But the order its problems had to be taken in, and which one turned out to be the real constraint rather than the loud one, is useful to somebody standing on different ground with a different problem.