When I look at a site, the first thing I want is the reason somebody else walked away from it. On anything worth having there is one, and it is rarely the price. Price is what people say when they do not want to explain the real objection.

This matters because the reason determines whether the site is an opportunity or a trap, and the two look identical from outside. A site that is hard because of something that can be worked through is an opportunity, and it is cheap precisely because the working through is uncertain. A site that is hard because of something immovable is a trap, and it is cheap for the same reason. Telling them apart before committing is most of the skill in this stage.

So the work is diagnostic rather than imaginative. I am not asking what could be built. I am asking what is actually preventing anything from being built, and whether that obstacle has a mechanism behind it or is simply a fact. Those require completely different responses and confusing them is expensive in both directions.

The second discipline is separating the site from the scheme. A developer who arrives with a building already in mind will find evidence the site supports it, because sites are ambiguous enough to support most readings if you want them to. I try to establish what the ground will permit before allowing myself an opinion about what should go on it, and I am not always successful at that.

The third is about time. Every difficult site has a duration attached to it, and the duration is not proportional to the difficulty. Some hard problems resolve quickly once the right question is asked. Others are hard in a way that simply takes years regardless of competence or resources. Misreading which kind you have is, in my experience, the most common serious error at this stage, more common than misreading the obstacle itself.

Here is where I will say something a lot of experienced people disagree with. I think most developers select sites too early in their own reasoning. The decision gets made on a first impression and everything after it is justification, including work that looks rigorous. The tell is that the analysis never changes the answer. If your diligence has never killed a site you were enthusiastic about, it is not diligence.

The counterargument is that conviction early is exactly what lets a developer move on a good site before somebody else does, and that hesitation has a cost that does not appear in any analysis. That is true and I have lost sites to it. I still think the error runs more often in the other direction.

The honest limit is that this stage is where I have been most wrong, and the wrongness is asymmetric in a way that hides it. A site I passed on that somebody else built well is a visible mistake I can learn from. A site I pursued that quietly consumed years is also a mistake, but it is much easier to narrate as bad luck, and I have certainly done that.

Where this crosses into law or land use, the diagnosis needs a professional rather than an opinion. What a site will permit is a question for a land use lawyer in that jurisdiction, and the honest answer is often that nobody can tell you yet.