Vol. 019: The Nod and the Shrug

June 10, 2026

Current Conditions

There is a conversation that happens in almost every land development meeting now. Someone says the old way of doing things is slow, wasteful, and expensive. Everyone nods. The diligence still takes a quarter, the entitlement still takes eighteen months, the surprises still arrive late. Then someone asks what tool the firm is actually using to change that, and the room gets quiet. The honest answer, more often than not, is some version of “we have Copilot in Excel sometimes.”

The interesting part is not the gap between belief and practice. Every industry has one. The interesting part is that the constraint people are describing has already lifted, and the practice has not noticed. For most of the history of this business, the limiting factor was not judgment. It was reading. The public record that governs an entitlement, the meeting minutes, the votes, the comprehensive plan, the fee schedules, the staff reports, was real and available and almost completely unreadable at human scale. So practitioners read a thin sample of it, layered institutional memory on top, and called the result experience. That was the correct strategy when reading was the bottleneck. It is worth asking whether reading is still the bottleneck.

Signals Observed

The clearest signal is not a survey of who uses what. It is that the same avoidable costs keep appearing in the public record, and they are avoidable specifically by reading.

Consider the shape of the failures this newsletter has documented across eighteen issues. A 931-unit subdivision in Wendell denied on traffic and school capacity, where the prior two years of hearings in that jurisdiction had been dominated by exactly those two themes. A 22x spread in townhome fees inside a single metro, discoverable from published schedules before a contract is signed and routinely discovered after. Text amendments that move the rules for projects not yet filed, visible in agendas weeks ahead and caught by almost no one. Two identical site plans, same builder, ten months apart in entitlement time, the difference sitting entirely in the political structure of the two jurisdictions. None of these were unknowable. Each was a reading failure dressed up as bad luck.

The second signal is that the information asymmetry has inverted, but pricing has not adjusted. The record today is genuinely machine-readable. Roughly 62,500 meetings across 200 municipalities in 24 states, something over two million searchable segments, sit in a structured form that did not exist a few years ago. A question like “has this board denied this product type in this district in the last three years” used to be a phone call to whoever had been around long enough to remember. It is now a query. Yet deals continue to be priced as though that question is still expensive to answer, which means the surprise, the denial nobody saw coming, has become the tell. A predictable outcome that still surprises the room is evidence that the reading did not happen.

The third signal is in what does get adopted. Where new tools have entered the workflow, they cluster around the horizontal and generic, the spreadsheet assistant, the general-purpose chatbot used to polish a memo. They stall at anything deal-shaped. The work product has not changed even where the inputs became readable. A diligence packet from 2026 looks like a diligence packet from 2016, which tells you the new capability is being used to type the old answers faster rather than to ask larger questions.

Forecast

If conditions hold, the gap does not close on its own, and it is more likely to widen than to narrow. The cost of reading the record fell sharply and roughly all at once. The habits built around scarce reading did not move with it, because the work is episodic. A firm doing four deals a year experiences the public record four times a year, which is not enough repetition to build a new muscle, and so each deal re-pays the manual cost as if it were the first. The likely result is a slow divergence: a small number of operators who have already changed how they read pricing risk more accurately and earlier, and a larger number who keep absorbing the avoidable costs and attributing them to a hard market. The divergence will be hard to see in any single deal and visible only in aggregate over a few years.

Watch Items

• Whether domain-specific tools, rather than horizontal assistants, reach the point where they fit a deal-shaped workflow rather than a calendar-shaped one
• Lender and equity diligence standards, which historically force adoption faster than internal preference does once a capital partner expects a question answered
• The verification problem, specifically whether practitioners develop trust in machine-read records the way they once learned to trust comparable sales data
• State zoning reform, which by making more approvals ministerial would shrink the political surface area that currently rewards deep reading
• Turnover, as the practitioners who hold the institutional memory retire and take the thin-sample method with them

Confidence Level

Medium-High. The cost side is well grounded in the public record and in this newsletter’s own back catalog; the adoption side is inferred from the persistence of avoidable, record-visible mistakes rather than measured directly, which is the appropriate place to hold some caution.

An aside. This issue was assembled, like the eighteen before it, by reading the public record at scale. That is offered not as a claim but as a data point: the reading is being done by someone. The only open question is whether it is being done before the deal or after.

Why This Matters

The decisions that determine whether a deal works are made early, before the contract, before the application, in the window where the only available information is the public record. For a long time that window was effectively dark, because the record could not be read fast enough to matter. It can be now. The firms that internalize that the bottleneck has moved will ask better questions earlier and price risk they used to discover late. The rest will keep nodding at the right moments and reaching for the spreadsheet. The gap between those two postures is small today and compounding.

Arc: The Reading Problem (Vol. 19 of 19-24). Entitlement work has always been a reading problem disguised as a relationship problem. The record was unreadable at human scale, so practice relied on a thin sample and institutional memory. That constraint has lifted. The cost now lives in the gap between what is readable and what anyone reads.