What changed
FACT (source: Federal Register 2026-11664): FEMA finalized new/modified Base Flood Elevations, base flood depths, SFHA boundaries, zone designations, and regulatory floodways via Letters of Map Revision, revising the effective FIRMs (and in some cases the FIS reports) for the listed communities. The notice states these determinations are finalized. INFERENCE: this is a recurring notice stream (the input itself notes ids 2710-2722 are further installments), not a one-time rule change.
Why now
HYPOTHESIS: nothing in the source document establishes a 'now'. This is a routine, continuously-published FEMA notice class, not a new mandate. The only genuine 'why now' is technical: national parcel coverage (Regrid-class datasets) and FEMA's NFHL/LOMC bulk data are now cheap enough for one person to run a nationwide polygon diff, which was not true a decade ago. That is a capability argument, not a demand argument, and it should be labeled as such.
Converging signals
WEAK. The input contains exactly one signal — a single FEMA notice — presented as its own demand evidence (linkage: 'structural: opportunity derived from this mandate'). There is no independent complaint, hiring, or spend signal in the input. Per the system's own lesson (conf 0.80) that Federal Register Notices are not filing mandates and should be filtered out relative to RULE/PRORULE, the convergence claim here is thin: the rule, the filer class, and the portal do not actually meet at a compelled filing.
Customer pain
FACT from the source: SFHA boundaries and zone designations change. INFERENCE (well-grounded in NFIP structure, not in this document): a parcel newly mapped into an SFHA triggers the federally regulated lender's mandatory-purchase requirement, so the owner must carry flood insurance. The owner's route out is an MT-1 package (Elevation Certificate, deed, plat, community acknowledgment) filed as a LOMA or LOMR-F. CRITICAL DISTINCTION the input gets wrong: the owner is compelled to BUY INSURANCE, not to FILE. The MT-1 is a voluntary, deadline-free opt-out. This is not a forced-filer class in the sense the founder's thesis means, and the source states plainly: 'DEADLINE: None stated.'
Who pays
Three candidate buyers, in descending order of reachability. (1) Licensed surveyors, PEs, and floodplain consultants — they must sign the Elevation Certificate, they already sell MT-1 work at roughly $500-2,500 per parcel (HYPOTHESIS: from general market knowledge, not from any provided source), and they currently find clients by word of mouth. A per-seat feed of 'these 380 parcels in your service area just entered the SFHA this month' is a lead product they can price against a single closed job. (2) Homeowners — highest emotional pain, worst unit economics: one-shot transaction, cold-mail acquisition, and an existing industry of 'flood zone removal' direct-mail firms already saturating them. (3) Lenders and servicers — they buy exactly this as 'life-of-loan flood zone determination' from entrenched national vendors under multi-year contracts. That channel is enterprise procurement and should be treated as closed.
Solved today
HYPOTHESIS (no source in input): FEMA notifies communities and publishes the LOMC database and NFHL; the Flood Map Changes Viewer and MSC let anyone look up a specific address. Surveying firms and specialist 'flood zone determination removal' consultancies do the MT-1 work manually, often on contingency against the avoided premium. Lenders buy automated zone-determination monitoring from national data vendors. Nobody in the input is documented as complaining, hiring, or paying.
Why current solutions are bad
The lookup is address-at-a-time and reactive: a surveyor cannot ask 'which parcels in my three counties changed zone this quarter?' without doing the geospatial diff himself. FEMA publishes the polygons but not the parcel join. That gap is real and is the only defensible technical wedge here.
Proposed product
A LOMR change-detection pipeline: poll FEMA's LOMC/NFHL publication stream, diff each revised SFHA polygon against licensed county parcel geometry, classify each affected parcel (newly in / newly out / zone reclassified / BFE changed), and expose it as (a) a territory-scoped alert subscription for surveying and floodplain-consulting firms and (b) a document-assembly workbench that pre-fills the MT-1 forms, pulls the deed and plat, and tracks the community acknowledgment form through signature. The founder never certifies anything; the licensed professional does. The software sells the lead and removes the clerical hours.
MVP version
Pick 3-5 coastal or riverine counties with recent LOMR activity and good open parcel data. Script the LOMC feed pull, run the polygon diff in PostGIS, and produce a weekly parcel-level change report. Hand-deliver it free to 20 surveying firms in those counties. Do not build the filing workbench until at least three firms say the list is worth paying for. Total build: a few weeks of one person's time plus a parcel data license.
30-day build
Validate the data spine, not the product. Confirm FEMA's LOMC/NFHL revision data is machine-ingestible at the cadence the notices imply and that revised polygons are actually retrievable per LOMR (this is the single biggest technical unknown and it can kill the idea outright). In parallel, call 25 surveying and floodplain-consulting firms and ask one question: how do you currently find MT-1 work, and what would a list of newly-affected parcels in your county be worth? Their answer decides everything downstream.
60-day build
If firms show interest, run the diff for their counties and deliver real lists. Instrument conversion: how many parcels on the list become a signed engagement. That ratio is the entire pricing argument. Begin the deed/plat retrieval integration only for counties where records are digitally accessible — this is unglamorous county-by-county plumbing and it is where the schedule will actually slip.
90-day revenue plan
Target 8-15 firms on a territory subscription in the $200-600/month range. That is roughly $2-9k MRR — a real business at solo scale, not a breakout. Per-filing revenue share on the workbench is a later, and legally messier, layer.
Distribution path
Direct outbound to a small, enumerable, non-enterprise buyer: state surveyor licensing rosters and ASFPM/state floodplain-manager association member lists make the entire buyer universe addressable by name. Demonstrated value works here — send the free county diff first, then charge. This suits the founder's stated selling style precisely.
Pricing hypothesis
Per-seat territory subscription, $200-600/month per firm per county cluster. Explicitly NOT the $300-800 per-filing fee the input proposes: the founder cannot sign an Elevation Certificate, so he cannot own the filing relationship or defend that price. Selling per-filing puts him in the homeowner-acquisition business, which is the worst channel in this idea.
Technical difficulty
Moderate and mostly logistical. Polygon diffing is standard PostGIS. The hard parts are: FEMA's revision data may not cleanly expose 'before' and 'after' polygons per LOMR (verify first); nationwide parcel geometry must be licensed and is uneven in quality; per-parcel mortgage data is NOT publicly available (deed and lien records are county-level and inconsistent), so the input's premise of diffing against 'mortgage data' is substantially harder than it sounds and should be dropped from the MVP.
Legal / regulatory risk
Low for the founder personally, with one sharp edge. He must not represent that he can get a property out of a flood zone, and he must not touch the professional certification. The 'flood zone removal' adjacent industry has a promotional-claims reputation problem; staying on the B2B side of it — selling data to the licensed professionals — avoids both the consumer-protection exposure and the unlicensed-practice question.
Platform dependency
None meaningful. FEMA cannot deplatform a tool that reads its public flood-map data. The real dependency is on commercial parcel data licensing, which is a vendor cost and a vendor risk, not a policy risk.
Founder fit
Genuinely good on the mechanics — public records, geospatial automation, small reachable buyer, sell-by-demonstration — and this is a real strength. But it is a weaker fit than the ELDT precedent, and the difference matters. In ELDT the regulation compelled the customer to file and the founder owned the submission. Here the regulation compels the customer to buy insurance; the filing is optional, deadline-free, and its critical path runs through a licensed professional the founder must recruit rather than replace. The founder's proven edge is monetizing compulsion. This idea monetizes motivation. Those are not the same asset.
Breakout potential
Bounded. The buyer count is the number of surveying and floodplain-consulting firms doing MT-1 work — thousands, not hundreds of thousands. The 5-million-NFIP-policy figure in the input is a population statistic, not a market: it describes people who might benefit, not buyers who will pay the founder. Treat it as noise. Realistic ceiling is a solid five-figure-MRR data business with 50-state replication, not a venture-scale outcome.
Final recommendation
CONDITIONAL — do not build as specified; run a two-week validation instead. The idea as framed (per-filing fee to homeowners, forced-buyer demand, mortgage-data diffing) rests on three claims the source does not support: that filers are compelled, that the founder can be the filer, and that mortgage data is available. Strip those away and a smaller, more honest business remains: a LOMR-to-parcel change feed sold to licensed surveyors and floodplain consultants. That version is solo-buildable, technically real, has a named and enumerable buyer, and fits the founder's demonstrated selling style. It is not the public-money forced-filer shape he is hunting, and it should not be scored as if it were — grading this as a FORCED BUYER opportunity would teach the system to promote discretionary opportunities dressed in mandate language, which is precisely the failure the Federal-Register-notice-filtering lesson (conf 0.80) was written to prevent. Two gates before any spend: (1) confirm FEMA's LOMC/NFHL data actually exposes per-LOMR before/after polygons — if it does not, the product does not exist; (2) get three surveying firms to state a price for the county change list. Fail either gate and kill it.
Next action
Spend one day on the data gate before spending one dollar on parcel licensing: pull FEMA's LOMC database and the NFHL for two of the communities named in notice 2026-11664 and determine whether the pre-revision and post-revision SFHA polygons are both machine-retrievable for a specific LOMR. Everything else in this brief is downstream of that single yes-or-no.