What changed
FACT (per the notices): FEMA publishes recurring Federal Register notices titled 'Changes in Flood Hazard Determinations' in which Letters of Map Revision (LOMR) finalize new or modified Base Flood Elevations, base flood depths, SFHA boundaries, zone designations and regulatory floodways for a table of named communities, revising the effective FIRMs and in some cases the FIS reports (2026-13898, 2026-13897, and eleven near-identical notices dated 2026-06-11 and 2026-07-09). INFERENCE: nothing in the statute or the portal changed; what is 'new' is only that the map revisions are continuous and machine-readable, and that the NFHL is published as an ArcGIS/WMS service that can be diffed programmatically. Be clear-eyed: this is a standing, decades-old process, not a fresh mandate.
Why now
HYPOTHESIS, not fact. The honest 'why now' is not a regulatory change — it is a tooling change: the NFHL is queryable as a service, LOMR effective dates are published on a predictable cadence, and parcel/geocoding data is cheap enough that a solo operator can compute 'which specific addresses just crossed into the SFHA' within hours of each notice. Nobody is compelled to start filing today who was not compelled last year. Anyone selling this on urgency is selling a story. The real timing argument is that Risk Rating 2.0 has made the premium delta from an SFHA designation salient to owners and lenders, and that no cheap, systematic, per-LOMR alerting product appears to occupy the niche — both of which are INFERENCES I have not verified against a source in this input.
Converging signals
Three things meet at one point: (1) a recurring federal notice that names exactly which communities' SFHA boundaries moved (FACT, cited); (2) a statutory forced action — the Flood Disaster Protection Act mandatory-purchase requirement means an owner with a federally-backed mortgage whose structure is newly inside the SFHA must buy flood insurance, and the only escape is a FEMA map-change filing (FACT as to the mechanism, per the notice text describing mandatory purchase); (3) a public, programmatically queryable geospatial layer (NFHL) plus a public submission portal (Online LOMC) that together make the diff-and-file loop automatable (FACT per the convergence description's portal section, which matches FEMA's published architecture). The convergence is real. Its commercial value is the part in doubt.
Customer pain
Two distinct pains, and they are NOT equally valuable. (a) HOMEOWNER: gets a letter, discovers a mandatory flood-insurance premium, has no idea a LOMA exists, and if they find out, faces a confusing MT-1 package requiring an Elevation Certificate. High acute pain, but a one-shot, low-trust, hard-to-reach buyer. (b) SURVEYOR / CIVIL-ENGINEERING FIRM: already sells Elevation Certificates and files MT-1/MT-2 packages, and its bottleneck is not the paperwork — it is *finding the parcels that just became addressable leads*, and then the clerical tedium of assembling and tracking each packet through the LOMC portal. This second pain is where money already changes hands. HYPOTHESIS: the clerical half of the surveyor's job is 1-3 hours per packet; I have no source in this input proving it.
Who pays
Primary: land surveyors, floodplain-consulting shops, and small civil-engineering firms that file MT-1/MT-2 packages — they pay per-seat or per-lead because each converted LOMA/LOMR-F is a $500-1,500 Elevation Certificate engagement for them. Secondary: flood-zone determination vendors and independent insurance agents who want a triggered list of newly-in-SFHA addresses in their book. The homeowner is the *beneficiary*, not the ideal buyer — selling to them is cold consumer marketing against a stranger's letter, which is exactly the profile the founder avoids. INFERENCE, flagged: no job posting, contract, or complaint in the supplied evidence proves any of these three parties currently pays for this specific service.
Solved today
Surveyors find LOMA work through word of mouth, realtor/lender referrals, and by manually watching FEMA's Map Service Center and the Flood Map Changes Viewer. Homeowners find out from a lender's forced-placement notice, then either eat the premium or hire a local surveyor. A cottage industry of 'flood zone removal' services already exists and markets directly to homeowners on a contingency or flat-fee basis. FEMA itself charges no fee for a LOMA. INFERENCE from domain knowledge; the notices in this input say nothing about any of it.
Why current solutions are bad
The LOMR notice tells you the *community*, not the *parcel*. The gap between 'Jefferson County's map changed' and 'these 412 specific addresses are newly inside the SFHA and 60 of them sit on natural high ground and will qualify for a LOMA' is a geospatial diff nobody is doing systematically and publishing. That gap is the entire product. Everything downstream — packet assembly, portal submission, status tracking — is commodity workflow software that adds margin but would not, alone, win a customer.
Proposed product
A LOMR change-detection engine with an attached filing workspace. Weekly (and on each Federal Register notice), pull the current NFHL, diff SFHA polygons against the prior snapshot for the affected communities, intersect the newly-inside geometry with county parcel/address data, and score each parcel for LOMA-qualification likelihood using publicly available elevation data (USGS 3DEP / county LiDAR) against the new BFE. Output: a ranked, geocoded lead list per county. Attached: an MT-1 workspace that pre-fills the property information form, tracks the community acknowledgment form, holds the surveyor's uploaded Elevation Certificate, and generates a submission-ready package for the Online LOMC portal. Sell the lead list as the wedge; sell the workspace as the retention.
MVP version
Pick 3-5 counties with good open parcel data and a recent LOMR. Build: (1) a Federal Register poller for 'Changes in Flood Hazard Determinations' that extracts the community table; (2) an NFHL snapshot-and-diff job producing newly-inside-SFHA polygons; (3) a parcel intersect + geocode step; (4) a LiDAR-vs-BFE qualification score; (5) a one-page PDF per county listing ranked candidate addresses. That is the whole MVP — no portal integration, no packet builder. Take that PDF to ten surveyors and see if any will pay $500/month for their county. If none will, the idea is dead and you have spent six weeks, not six months.
30-day build
Do not build the workspace. Build the diff. Week 1-2: NFHL snapshot pipeline + LOMR notice parser, backfilled over the eleven June/July 2026 notices already in hand. Week 3: parcel intersect for three counties; buy parcel data if the county does not publish it (Regrid or equivalent — modest, fundable spend). Week 4: hand-verify 30 parcels against the FEMA Map Service Center to measure your false-positive rate. Simultaneously: call 20 surveying firms in those counties and ask what they pay per lead today. If you cannot get a number out of any of them, that is your kill signal.
60-day build
Add the LiDAR/BFE qualification score and measure it against a sample of *granted* LOMAs (FEMA publishes LOMC determinations — use them as labels). A defensible qualification model is the only piece of this that an incumbent cannot copy in a weekend. Sign 3-5 design partners at $300-750/month per county with an explicit promise: 'we will show you every parcel that entered the SFHA in your county within 72 hours of the LOMR.' Begin the MT-1 workspace only against a paying partner's actual workflow.
90-day revenue plan
5-10 county subscriptions at $500/month = $2,500-5,000 MRR. Layer per-packet fees ($75-150 to the surveyor for an assembled, portal-ready MT-1, not the $150-400 direct-to-homeowner figure in the input — the homeowner channel is a distraction). Realistic first dollar: day 60-90 from a design partner; realistic durable revenue: month 5-6. The 30-day-revenue framing does not apply here and pretending otherwise would be dishonest.
Distribution path
This is the weakest dimension and the most likely cause of failure. There is no marketplace, no app store, no viral loop. The channel is: state surveyor and floodplain-manager associations (NSPS chapters, ASFPM state chapters, state boards of licensure), which have conferences, newsletters, and email lists a solo operator can reach; plus direct outbound to the 20-50 firms per state that actually file map-change requests. The founder sells through demonstrated value, and this product demos itself — walk into a surveyor's office with a list of 400 addresses in his county that just entered the SFHA and the demo is over. But that is a slow, manual, county-by-county grind, not a self-serve funnel. Do NOT attempt homeowner direct mail: cold consumer acquisition against a scary FEMA letter is adjacent to the reputation of the worst actors in this space, and the CAC will not clear a one-time $300 fee.
Pricing hypothesis
$500/month per county for the lead feed (annual discount), $150 per assembled MT-1 packet, $1,500/month for multi-county firms. Rejected: the input's $150-400 homeowner per-filing fee. It looks attractive because the owner's avoided premium is large, but it puts you in a one-shot consumer transaction with a customer who found you by cold outreach, and it puts you uncomfortably close to representing a property owner before a federal agency on a technical determination you cannot certify.
Technical difficulty
Moderate, and honestly assessed: the NFHL diff is real GIS work (polygon topology, coordinate systems, versioning of a layer that FEMA republishes rather than versions cleanly), parcel data is fragmented across 3,000+ counties with wildly different licensing, and the LOMA-qualification score requires reconciling LiDAR-derived lowest-adjacent-grade against a BFE surface — which is precisely the judgment a licensed surveyor is paid to certify. The founder's systems-thinking and public-records strengths map well onto the first two. The third is where a naive model produces confidently wrong answers and burns the surveyor's trust on his first month. Budget for the qualification model to be the hard part; treat everything else as plumbing.
Legal / regulatory risk
Real and specific, and the standing instruction not to flag heavy_compliance does not fully absolve it here. A LOMA determination rests on an Elevation Certificate signed by a licensed land surveyor, professional engineer, or architect. If the product tells a homeowner 'you qualify' and sells them a filing, that edges toward the unlicensed practice of surveying in states that define it broadly, and toward FEMA's own guidance that MT-1 elevation data must be certified. Selling to licensed surveyors as a lead-and-workflow tool sidesteps this cleanly — the licensee makes the determination, signs it, and owns the liability. That is not a nicety; it is the reason to choose the B2B channel over the consumer one. Second risk: accuracy claims. A lead list that says 'newly in SFHA' about a parcel that is not exposes you to a professional's complaint. Ship it with a stated false-positive rate and a 'verify against the effective FIRM' disclaimer. Third, minor: parcel data licensing — several county and commercial parcel sources prohibit redistribution, and your product output is arguably a derivative.
Platform dependency
Low, correctly. FEMA cannot deplatform you; the NFHL and the Online LOMC portal are public infrastructure and the Federal Register is a public API. The genuine dependency is data-shape stability: if FEMA changes the NFHL publication schema or the notice table format, the pipeline breaks — that is a maintenance cost, not an existential platform risk. One asymmetric tail risk worth naming: the NFIP requires periodic congressional reauthorization and has lapsed before. A prolonged lapse suspends new policy issuance and would freeze the urgency that drives LOMA filings. Low probability of a durable lapse; nonzero.
Founder fit
High, with one caveat. The shape is exactly the ELDT/Training-Provider-Registry pattern he has already shipped: read a federal mandate, identify the compelled class, build the submission layer against a government portal, charge per filing. Public-records extraction, geospatial diffing of government data, and per-transaction pricing against a portal are all in his demonstrated repertoire. The caveat is that ELDT had a captive, self-identifying buyer (the training provider *knows* it must file) while here the compelled party (the homeowner) does not know he has an option, and the buyer who does know (the surveyor) is a professional-services firm the founder must reach through associations and cold calls. That is a harder distribution motion than ELDT's, and it is the difference between a 9 and an 8 on fit.
Breakout potential
Bounded but real. The same NFHL-diff engine underwrites adjacent products with no new data work: a triggered feed for independent insurance agents ('these 40 addresses in your book just entered the SFHA'), a community-official dashboard for the floodplain administrators the notices say must administer the revised maps, and a lender/servicer compliance check. Nationwide, this is thousands of surveying firms and every LOMR cycle regenerates inventory — no churn cliff. It is not a venture-scale business and should not be built like one. Ceiling: a $500k-2M/yr owner-operated data-and-workflow company. That is a very good outcome for this founder and a bad one for anyone who mistook it for something larger.
Final recommendation
CONDITIONAL BUILD — reframe, then de-risk cheaply before committing. The idea as titled (auto-assemble MT-1 packs, charge homeowners per filing) should be rejected: it sells the cheap half of a surveyor-gated deliverable to a cold consumer, and it walks toward unlicensed-practice exposure. The idea that survives is narrower and better: **be the system of record for what changed in the National Flood Hazard Layer, and sell that to the licensed professionals who monetize the change.** That version has a reachable buyer, a demonstrable-in-thirty-seconds value proposition, a per-county recurring price, and a technical moat (the qualification model) if and only if the model actually works. Founder fit is genuinely high — this is the ELDT shape with a harder distribution motion. But two things must be true and neither is established by the evidence supplied: (1) a surveyor will pay for the lead list, and (2) LiDAR-vs-BFE scoring beats eyeballing the map on marginal parcels. Both are answerable in under six weeks for under $5k. Spend that, not six months. Score it B, not A: the forced-buyer structure is real, the mandate is fifty years old, and the demand evidence in this input is a single notice echoed twenty times.
Next action
Before writing any code: call fifteen land-surveying and floodplain-consulting firms in three counties named in the 2026-06-11 and 2026-07-09 LOMR notices. Ask exactly two questions — 'how do you currently find out which parcels entered the SFHA after a LOMR?' and 'what would you pay for a verified list of them within 72 hours?' If three or more name a dollar figure at or above $300/month, build the 30-day diff MVP. If none does, kill it and do not spend the six weeks.