For a long time my work has concentrated on a fairly narrow band of problems, and the consistency strikes me more now than it did at the time. Putting sites together. Getting them entitled. Moving development capacity onto them. Getting preservation approvals where a designation was involved, and negotiating what a project would give back. Two decades on, that band has barely widened, while the buildings coming out of it changed beyond recognition.

What ties those together is that they all sit before construction. That is where most of what I have done has actually happened. By the time a site is ready to build, the decisions that determine what the building will be have largely been taken, and the years of work that took them are invisible in the finished thing. I find that genuinely satisfying and I understand why it makes for poor storytelling.

The buildings themselves have gone in several directions at once. Ground up development and adaptive reuse. Landmark restoration. Residential of different kinds, mixed use, hospitality, branded residences. I am occasionally asked what the house style is, and the honest answer is that there is not one. Letting each site produce its own answer is a position rather than an absence of one, and I would defend it.

What they share is difficulty arriving from four directions at once: from the ground itself, from whoever has to approve it, from the engineering, and from how it gets paid for. Trouble in one of those is an ordinary project. The ones I remember had trouble in three at the same time, which is a different animal, because they compound each other instead of queueing up politely.

The recurring pattern has been to buy something everyone else had written off and then find room inside it the earlier reading missed. That is far less clever than it sounds. It is slow work through records and rules, it yields nothing for a long time, and then it yields a great deal at once.

It has also covered the whole life of a development rather than one slice of it. Buying and financing at one end; getting a project entitled; building it, or restoring it, or converting it, which are three different trades wearing one word; and then letting or selling it at the other end. Holding the construction capability close to the development one is what has made it possible to stay on a project from the first look at a site through to handing it over.

I write about all of this now with more interest than I did at the beginning, which I think is a function of having enough completed work behind me to see which decisions actually mattered. Quite a few of the ones that felt decisive at the time did not, and several that felt procedural turned out to determine the outcome.

The firm itself was arranged at the start so that acquisition, development and construction sat together rather than being contracted between separate parties. I have written about what that actually buys elsewhere and will not restate the argument here. What matters for this account is that it was decided early, before there was much of a firm to arrange, and it has not been revisited since.

More than twenty years on I still run it. The scale has grown a great deal since 2002 and the reach has widened, and the kind of problem that interests me has not moved at all.