A developer we know puts a parcel under contract and starts a clock. Sixty days, sometimes ninety, to decide whether the deal is real.
In that window, a small number of people read an enormous number of pages. The title commitment and every exception document attached to it. The ALTA survey. The wetlands delineation. The Phase I. The geotech report. The zoning ordinance, and then the overlay district, and then the three amendments to the overlay district. The utility will-serve letters, or the absence of them. Prior plats. Easements recorded in 1974 by a company that no longer exists. HOA declarations. The county's capital facilities plan.
Somewhere in that stack is the thing that kills the deal. It's usually one sentence. A setback that makes the density model fall apart. An access easement that runs the wrong direction. A sewer line that isn't where the seller's broker said it was.
The whole discipline of land acquisition is the art of finding that sentence before you go hard on your earnest money.
Where the time actually goes
Here's the uncomfortable audit. In a typical diligence cycle, the majority of the hours are not spent on judgment. They're spent on retrieval.
Someone reads 400 pages looking for whether there's a conservation easement. Someone else reads the same 400 pages looking for utility capacity. A third person reads them again to build the summary memo for the investment committee. Each of those passes is a person, at a high loaded rate, performing the cognitive equivalent of Ctrl-F on paper.
Price it. A land acquisition analyst at $85,000 all-in is running about $50 an hour loaded. Say twenty-five hours of pure document review per deal — conservative. That's $1,250 a deal in reading. Do fifteen deals a year to close four, and you're at nearly $19,000 annually in reading, most of it on deals you walked away from. Add the principal's time reviewing the memo, and the two weeks of calendar drift while everyone waits for someone else to finish reading, and you've got a number that's real.
But the dollar cost isn't the interesting part. The interesting part is the deals you didn't look at.
The real cost is opportunity, not labor
If diligence takes six weeks of somebody's undivided attention, you can only run so many deals in parallel. That's not a staffing problem you can hire your way out of at a fifteen-person shop — every additional analyst is a real, permanent cost, and land teams are famously lumpy.
So you triage. You pass on the marginal parcel because you don't have the bandwidth to underwrite it properly, and you tell yourself it probably wasn't a great deal anyway. Sometimes that's true. Sometimes the parcel you passed on becomes somebody else's best project of the decade.
That's the actual cost of slow diligence: deal flow you never processed. It doesn't appear anywhere in your financials. It's the most expensive line item you don't have.
What a machine does well here, and what it doesn't
Reading is a machine task. Judgment is not. The whole design of a good system here is drawing that line honestly.
A machine can ingest the full document set and answer questions against it with citations — what are the setback requirements, and where does it say that; list every easement burdening the property with recording dates and grantees; does the geotech report mention expansive soils, and on which page. It can flag the exceptions in a title commitment that historically matter and the ones that don't. It can extract every date, every acreage figure, every reference to a utility, and lay them next to what the seller represented so the discrepancies surface immediately. It can produce a first-draft diligence summary in an hour that would have taken an analyst two days.
What it cannot do — and this is not a temporary limitation, this is the shape of the thing — is decide whether the county planner will actually approve the rezone. That's a judgment made of relationships, political read, and twenty years of watching how this particular commission behaves in an election year. It cannot decide whether the price is right given where you think rates are going. It cannot walk the site and notice that the drainage is wrong in a way no report captured.
That's the deal-maker's job. That's the whole job, actually. And right now that person is spending three days a week reading PDFs.
What changes when you split the work correctly
The analyst's role stops being retrieval and becomes verification. The machine produces the extraction with citations; the human checks the ones that matter and applies judgment to what it found. That's a fundamentally different — and better — use of a smart person.
The calendar compresses. Not because anyone works faster, but because the sequential reading passes collapse into one ingestion and many queries.
And the funnel widens. If preliminary diligence on a parcel takes four hours instead of four days, you can afford to look at parcels you would previously have passed on without opening the file. You'll kill most of them faster, too — which is the underrated half. Knowing in a day that a deal is dead is worth nearly as much as knowing in a day that it's alive.
The point
The reason to build this isn't to have fewer analysts. It's to have analysts who spend their time analyzing, and principals who spend their time in front of sellers and municipalities and capital partners instead of in front of a 300-page PDF at eleven at night.
The parcel doesn't care how hard you read. It cares whether you found the sentence.