Twelve years getting land ready to build on.

A carrier decides it needs coverage in a particular place and draws a circle. That circle is the whole brief. Somewhere inside it there has to be a piece of ground whose owner will lease, whose title is clean, whose soil will hold a tower, whose jurisdiction will permit one, and which is close enough to power and fibre to be worth building. Nobody knows which parcel that is, and roughly six candidates get researched for every one that gets built.

The work is finding out, in an order that spends the least money before the first thing that can kill the site kills it. That order is the craft. Ground is cheap to evaluate and expensive to survey. Zoning feasibility is a phone call and a code section; a variance is nine months. So the questions that can end it get asked first, and the expensive work only starts on ground that has survived them.

Then the parallel phase. Survey, title, environmental, geotechnical, and the utility and structural coordination all run at once, against one schedule, because running them in sequence adds a year nobody has. Each comes back correct on its own terms, and the failures are almost always in the seams: a survey that assumes an easement footprint the title work has not confirmed, a geotechnical report that answers a question about a tower position the design has since moved.

Then the public part. An application, a staff review, and a hearing in front of a body that can say no in a room open to the neighbours. That is the only step that cannot be quietly redone, and it is the one everything upstream was protecting.

Then the handoff, which is the part that determines whether any of it was worth doing. Construction gets a package, or construction gets a phone call, and a project that lives in one person’s head has to be rediscovered by whoever inherits it, usually under schedule pressure.

A carrier ran a programme to get off its most expensive lease sites. The brief was a leasing exercise: find cheaper colocations and move onto them. That is a sensible instruction and it was written by people who knew what they were doing.

Working the first tranche, I found that several of the jurisdictions imposed no zoning or permitting requirement at all. Where that is true the arithmetic inverts. The cheapest long-term position is not a cheaper lease. It is the client owning the asset, because the barrier that normally makes building slower and more expensive than renting simply was not there.

So I took it back to them rather than executing the programme as written. They agreed, and the scope changed from relocating onto other people’s towers to building the client’s own.

The transferable part is not the answer, it is where the answer came from. Nobody had read the ordinances before the programme was scoped, because at scoping time it looked like a real-estate question. It was a zoning question wearing a real-estate question’s clothes, and the only way to find that out was to read the code for the specific jurisdictions rather than reason about the category.

The archive holds 89 presentation decks I built myself, including the variance and special-use presentations taken to city councils and boards. That is what “I build the materials I present” means in file terms.

It also holds the reference library the team ran on rather than asked about: state regulatory folders for eleven states, the raw land and colocation workflows, a permitting notes standard, a ground lease extension amendment template, and separate photo-log standards for candidate, colocation and rooftop sites.

On the method: these are counts of files whose names match a document type, not a hand audit of 16,874 documents. They are honest about the shape of the work and they are not exact. Where this site gives an exact figure, it says so, and this is not one of those.