What changed
FACT (federalregister.gov, 2026-06-11 and 2026-07-09): FEMA published multiple notices requesting comments on PROPOSED flood hazard determinations — additions or modifications of Base Flood Elevations, base flood depths, SFHA boundaries or zone designations, and regulatory floodways on preliminary FIRMs and FIS reports for listed communities. FACT (same notices): the FIRM and FIS report are the basis of the floodplain management measures a community must adopt or show evidence of to qualify or remain qualified for the NFIP. INFERENCE: nothing about this process is new — FEMA has issued these notices on a rolling basis for decades. What is new is only that this particular batch of communities is now in-window.
Why now
HYPOTHESIS, not fact: the input contains no evidence of a rule change, a new appropriation, or a new mandate. The 'why now' is structural rather than temporal — at any given moment some set of communities is inside an open comment/appeal window, and the input shows two notice batches roughly 30 days apart (2026-06-11, 2026-07-09), consistent with a continuous cadence. A founder should be honest that this is an evergreen, always-open opportunity, not a closing window. That cuts both ways: there is no urgency-driven land-grab, but there is also no expiry.
Converging signals
Three things do meet at one point: (1) a federal notice that names specific communities and opens a comment period; (2) a statutory appeal right for owners and lessees, exercisable only during that period and only with supporting scientific or technical data; (3) a professional class (licensed civil engineers, hydrologists, land surveyors) who alone can produce that data. The convergence is thin, however, because all three have coexisted for thirty years without a software layer emerging — which is evidence that the gap is not as unexploited as it looks.
Customer pain
INFERENCE (not evidenced in the input): a property owner or developer discovers a new BFE or SFHA designation only when their lender demands flood insurance at closing, by which point the appeal window has closed and the only remedy is the slower, costlier LOMA/LOMR-F process. Engineering firms lose appeal work they never knew was available. No complaint threads, forum posts, or job ads validating this pain appear in the provided evidence — this is asserted, not proven, and must be validated by customer interviews before any build.
Who pays
Ranked by realism: (1) civil-engineering and land-surveying firms with floodplain practices — they bill hourly for the H&H study an appeal requires, so an alert that surfaces qualified work is a lead-gen product, and lead-gen for professional services has a well-established price ceiling; (2) production homebuilders and land developers holding parcel portfolios in growth counties; (3) HOAs and large commercial property owners; (4) NOT the communities themselves — that is municipal procurement and should be avoided.
Solved today
INFERENCE: floodplain consultants and engineering firms subscribe to the Federal Register themselves, or learn from the community's own outreach (FEMA requires community notification and a local newspaper publication). FEMA's Flood Map Changes Viewer and the Map Service Center already publish preliminary FIRM data publicly and free. Some firms maintain internal watch lists. Regional consultants who specialise in map revisions market directly to developers.
Why current solutions are bad
INFERENCE: the Federal Register notices list communities by name and CID, not by parcel, so mapping 'is my property affected' requires pulling the preliminary FIRM geodata and intersecting it with a parcel boundary — a step no one does at scale for third parties. That geospatial join is the only genuine technical work product here. Everything else in the chain is a PDF and an email.
Proposed product
A monitor, not a filing tool. Ingest FEMA's proposed/final/changes flood hazard determination notices from the Federal Register API; extract the community and CID table; pull the corresponding preliminary FIRM geospatial layers from FEMA's public map services; intersect the new SFHA and BFE polygons against a customer-uploaded parcel list (APN or geometry); emit a dated alert stating which parcels changed zone, by how much, and the last day of the comment period. Sell the alert. Sell a per-appeal 'evidence package' (parcel map, before/after zone, elevation delta, the exact notice citation, the community CEO contact) that a licensed engineer stamps and files. The founder does NOT file the appeal and must not imply he can.
MVP version
Federal Register API poller filtered to FEMA flood hazard determination notices; a parser for the community/CID table; a PostGIS load of FEMA NFHL and preliminary FIRM layers; a parcel-upload endpoint accepting APN lists or GeoJSON; a nightly intersect job; an email/webhook alert with a per-parcel before/after zone diff and the window close date. No portal integration exists to build — appeals route on paper through the community Chief Executive Officer to the FEMA Regional office, so there is no submission API to automate. This is the single most important difference from the ELDT product.
30-day build
Do not write the geospatial pipeline yet. Call twenty floodplain-practice engineering firms and ten regional land developers. Ask one question: 'when a preliminary FIRM drops on a county you work in, how do you find out, and what does missing it cost you?' Simultaneously, pull the last twenty-four months of FEMA proposed determination notices and count how many appeals were actually filed against them (FEMA regional offices and the LOMR docket will show this). If the true nationwide appeal volume is in the low hundreds per year, the packaging business is dead and only the alerting business survives. Build the Federal Register ingest — it is two days of work — so the interviews are backed by a live list of open windows.
60-day build
Assuming interviews clear the bar: build the parcel intersect. Ship a free public 'is my county in an open window' page seeded with every currently-open notice, gated on email. Hand-deliver the first ten alerts as PDFs, manually produced, to named engineering firms — no product, no login. Charge nothing. Measure whether any of them converts the alert into billable work; that conversion, not signups, is the only number that matters.
90-day revenue plan
Convert the hand-delivered firms to a per-seat subscription at $249-$499/month for unlimited county watches, and offer developers a per-parcel-portfolio monitoring tier. Ten seats at $349 is $3,490/month recurring — a real but small business. The per-appeal package fee ($1,500-$3,000) is upside, not the base case, and should be assumed to be zero until an actual appeal is packaged and paid for.
Distribution path
ASFPM (Association of State Floodplain Managers) chapter meetings and their state chapter mailing lists; the Certified Floodplain Manager community; direct outreach to the fifty largest firms with floodplain practices; a free open-windows page as SEO bait for 'preliminary FIRM [county]'. This is a small, list-addressable, conference-attending professional community — reachable without ad spend, which fits the founder. It is also a relationship-driven community, which does not.
Pricing hypothesis
$349/month per seat for firms; $99/month for a developer monitoring up to 500 parcels; $1,500 per appeal evidence package. Free county-watch tier as the funnel.
Technical difficulty
Low-to-moderate. The Federal Register API is documented, free, and stable. FEMA's NFHL and preliminary map data are public. The parcel intersect is standard PostGIS. The genuine difficulty is that parcel boundary data is fragmented across 3,000+ county assessors, so the founder must either buy a national parcel layer (Regrid, ~$5-30k/yr — affordable given stated runway) or require customers to upload their own geometry. Requiring upload is the right MVP choice and eliminates the cost.
Legal / regulatory risk
Moderate and specific. Producing a 'flood zone determination' or an elevation analysis for a third party edges toward the practice of engineering and land surveying, which is licensed in every state. The founder is not licensed. The product must be positioned strictly as a notification and document-assembly service, with all technical certification performed by the customer's own licensed professional, and with explicit disclaimers. This is a real constraint, not a formality: it caps the product at alerting and packaging and forecloses the higher-margin analysis work. Separately, an alert that is late or wrong could be alleged to have caused a missed appeal — errors-and-omissions coverage and a hard liability cap in the terms of service are mandatory.
Platform dependency
Low. The Federal Register API and FEMA map services are public government systems with no owner who can deplatform the product. No app store, no marketplace review. This is the strongest structural feature of the idea.
Founder fit
Good but not the bullseye. It matches his public-records scraping, geospatial-adjacent systems work, and preference for compliance monitors. It does NOT match the ELDT shape that made him money: in ELDT a federal rule compelled a defined class to submit through a portal, and he built the submission layer and charged per submission. Here, the compulsion falls on the COMMUNITY (adopt the maps or lose NFIP eligibility) while the APPEAL — the thing he would monetise — is entirely OPTIONAL, filed by a self-selected minority of owners, on paper, through a human CEO, with no portal to integrate. He would be selling a discretionary lead-gen alert to professional-services firms, not a mandatory filing tool to forced filers. That is a materially worse business, and the brief should not let the FORCED BUYER label on the evidence obscure it.
Breakout potential
Modest. The parcel-monitoring engine generalises: the same intersect-a-parcel-list-against-a-changing-government-layer pattern covers wetlands delineations, EPA NPL site boundaries, zoning overlay changes, DOT right-of-way takings, and state-level floodplain rulemaking. If the FEMA wedge works, 'government geospatial change monitoring for parcel portfolios' is a genuinely larger and more defensible product. That optionality is the strongest argument for building the first version.
Final recommendation
CONDITIONAL — and the condition should be treated as likely to fail. Do not build this as the appeal-packaging business the convergence describes; that business rests on an unmeasured appeal volume and an optional filing, and it fails the founder's own forced-filer test. There is a narrower idea worth two weeks of unpaid diligence: a monitoring product that intersects FEMA's changing flood layers against a customer's parcel list, sold to engineering firms and developers as lead-gen and portfolio risk. Before writing the geospatial pipeline, spend thirty days doing two things — count actual appeals filed against the last twenty-four months of proposed determinations, and interview twenty floodplain-practice firms about how they currently learn of preliminary FIRMs. If appeal volume is under roughly 500 per year nationally, or if the firms report they already track this adequately, kill it. If both come back favourable, the parcel-change-monitoring engine is worth building, primarily because it generalises beyond flood to wetlands, zoning, and right-of-way — that generalisation, not the FEMA appeal, is the actual business. Ranked against the founder's ELDT template, this is a clear step down: no portal to integrate, no compelled submission, no per-transaction hook. He should keep hunting for a PRORULE-class mandate with a named filer class and an electronic submission system, and treat this as a fallback.
Next action
Query the Federal Register API for all FEMA 'Proposed Flood Hazard Determinations' notices published in the last 24 months, extract the community/CID tables, then cross-reference against FEMA's LOMR and appeal records to produce a single number: how many BFE appeals were actually filed. That number, obtainable in under a day from public data, decides whether anything here is worth building. If it is small, stop.