License one Project workspace. Reach owners lawfully. Keep the legal economics.

The prize behind A0. Condemnors engineer the alignment, enumerate the tracts, and order title and appraisal work before the first owner sees a number; firms meet the same Project one inbound call at a time. At case-model defaults that leaves ~96% of the path unrepresented — roughly $1B in net fees on the golden window alone (~$90K per represented Parcel) that no one collects. The volume is contingency-financed: the clients already exist as affected owners, and no retainer stands between a firm and the work. Every dollar of it is firm money — Allodial's wall means we never touch the fee. The mission is to hand firms the board the condemnor already has.

The wall that keeps Allodial out of legal fees also caps our per-matter upside; a firm's contingency is uncapped. Chasing a bigger share now would mean pricing against the fee, which the architecture forbids — so the lever is the other axis entirely. Win on activation: the product that gets more of the path represented, and then makes sustaining that level cheap, is worth more to a firm every year it runs, and a product worth more can be priced higher without ever touching a fee. Volume first buys the proof and the install base; the bigger slice is earned later, from a product that demonstrably raised the number.

The proposition is one-sided in the owner's favor: contingency representation, no upfront cost, and challengers averaging well above the first number. What has been missing is outreach, not motive. Two things stand in the way, in order. Awareness and access to these people is the majority of the battle — knowing who they are, reaching them at all, and being findable when they go looking. Then the next challenge is feeling known: an owner who receives a letter about their own land, their own Project, their own milestones responds differently than one who receives a form. A tailored letter, email, inbound presence, and clear search are the system we build to make that close. The channel machine's ≈3% illustrative funnel is one generic wave; the claim is that top-tier craft, multiple firms, multiple waves, and a standing presence across the whole window compound well past it — which is what A0's 20% start actually requires.

The records are public and that is the trap: assembling route geometry against parcel fabric against owner identity, grading every join, proving each claim against its source, and keeping the whole thing current as sources rot is a standing engineering effort, not a project. 33 live object kinds, 169,438 cells over 31,709 nodes, 4,021 distinct public sources read, and a grading loop that refuted 1,148 of its own claims outright. A four-person practice does not run that. Neither does a shop that scrapes it once and ships a spreadsheet — the difficulty is not the getting, it is making thousands of unlike records agree and keeping them agreeing.