Off-market site flow, built from public data.

The question I started with

Before writing a line of code I spent the first stretch on the industry rather than the data, because the fastest way to build the wrong thing is to automate a process nobody actually uses. Two things came out of that.

First, the tooling gap is real and it is specific. The platforms that sell themselves as land-sourcing tools — the parcel aggregators, the ownership-data products, the site-selection suites — overwhelmingly filter on parcel size, zoning designation, and owner record. Those are the easy layers. They are easy because they come prepackaged statewide. Every developer running those tools is looking at the same filtered list, which is exactly why the good sites are competitive by the time they surface.

Second, that is not what actually kills a deal. Talking through how these sites die, the pattern is that a parcel clears acreage and zoning, gets a letter of intent, and then fails on something physical: a floodplain nobody looked at, a wetland that shrinks the buildable footprint below the unit count, or a sewer connection that turns out to be 2,000 feet and a lift station away. Those are the constraints that determine whether a site exists, and they are the ones the commercial tools do not filter on.

So I built the screen around the constraints that kill deals rather than the ones that are convenient to query.

How it was assembled

Five counties — Wake, Durham, Johnston, Orange and Chatham — and four public data sources, none of them purchased: the NC OneMap statewide parcel layer, the published gravity sewer networks from Raleigh Water and the Town of Cary, the FEMA National Flood Hazard Layer, and the USFWS National Wetlands Inventory. Everything was pulled fresh on August 6.

The density model is not a rule of thumb. It is calibrated against Litchford 315: 240 units on 14.3 acres, which is your own executed proof that the math works at that density. Each parcel is reduced by its floodplain and wetland area, then by a site-loss factor for roads, stormwater and setbacks, and the remainder is run at the Litchford ratio for three and four stories.

How 18,098 became 564

The under-improvement filter is doing quieter work worth calling out. Screening for parcels whose improvements are worth 30% or less of total assessed value is a proxy for an owner whose land is worth more than what is standing on it. That is a seller who has a reason to take a call, which is a different question from whether the site is buildable, and the reason the outreach list is 40 rather than 564.

Three of the forty

Owners, parcel IDs and addresses are withheld on this page. Everything else is exactly what the screen produces per site.

Each of the 40 also carries the owner of record, their mailing address, ownership tenure, and the source layer and retrieval date behind every number, so verifying a site takes half a minute rather than half a day.

What it does not know yet

Worth being straight about, because this is the distance between a list and a pipeline.

• Sewer capacity, as distinct from sewer proximity. A main at the property line means nothing if the basin is allocated out. Charlotte has already suspended capacity review in five basins. This is the highest-value thing to add and part of it will always be a phone call.
• Zoning. North Carolina publishes no statewide layer, so this is assembled jurisdiction by jurisdiction. It is the largest remaining piece of work and the one your team's knowledge shortcuts most.
• ETJ. An unincorporated parcel is often governed by a city's rules rather than the county's. Getting that wrong is the quietest way to lose a month.
• Riparian buffers. Nearly all of this footprint sits in the Neuse basin, at 50 feet with 30 of it no-disturb.
• Coverage. Only Raleigh Water and the Town of Cary publish their gravity mains. Durham, Apex, Holly Springs, Chapel Hill and Fuquay-Varina are not in yet, which is why Wake dominates the result. That is a gap in the data, not a judgment about those markets.

Why the timing matters

Two North Carolina statutes move the buy-box inside the next five months. Parking minimums end statewide on January 1, 2027. And the 2026 Regulatory Reform Act, passed August 4 and awaiting signature, would permit residential redevelopment of previously developed commercial land by right in Raleigh, Cary and Charlotte, with no local height cap below 60 feet.

Meanwhile multifamily starts nationally are down 73% from peak, the lowest since 2011, with deliveries falling through 2027. A site controlled now and entitled through next year opens into the thinnest supply in a decade. Anyone screening against today's rules is screening against rules that expire.

What this is actually an example of

Site sourcing is one bottleneck. It is the one I could see from the outside, which is the only reason it is the one I built.

The useful thing here is not the list. It is that a constraint your team carries in its head — where the sewer is, what the flood map does to a footprint, which owner has a reason to sell — turned out to be expressible as a filter, and that filter now runs across 18,000 parcels in the time it takes to read this page. That same move applies to most of the repetitive judgment inside a development shop. Some candidates:

I do not know which of those is the expensive one for Dewitt. That is exactly what the next conversation is for.

The next step

A Discovery Session. Half a day with you and whoever else owns the operational reality, structured in two halves.

The first half is this build. We pressure-test the filter against how your team actually thinks, correct what it gets wrong, and walk the sites that survived, with owners and addresses attached.

The second half is wider. We map where the repetitive, judgment-heavy work actually sits across your operation, and rank it honestly by what it costs you now against what it would take to fix. Some of it will be worth building. Some of it will not be, and you will get told which is which.

You leave with your screening criteria written down in your own language, the ranked Triangle list, and a prioritized map of the operational opportunities worth pursuing, whether or not you pursue them with me.