What changed
FACT (from source): DOL's Employment and Training Administration issued an NPRM, 'Improving Wage Protections for the Temporary and Permanent Employment of Certain Foreign Nationals in the United States' (published 2026-03-27), proposing to amend the regulations governing the PERM permanent labor certification program and Labor Condition Applications for H-1B, H-1B1, and E-3 visas. The specific change is to the computation of wage levels under the Department's four-tiered prevailing wage structure derived from the BLS Occupational Employment and Wage Statistics (OEWS) survey. DOL states the aim is to better align prevailing wage levels with wages paid to similarly employed U.S. workers and to strengthen program integrity. FACT: this is a Notice of Proposed Rulemaking soliciting comment — it is not a final rule, and the source text does not state a comment-close date or an effective date.
Why now
HYPOTHESIS: if finalized, a change to the tier formula does not merely affect new filings — it re-prices every SOC-code-plus-area combination in the system at once, meaning an employer's entire existing sponsored population must be re-evaluated against the new Level 1-4 breakpoints. The window to be the tool that already has the answer is the comment period and the gap between final rule and effective date. FACT: the source text confirms the four-tier structure and the OEWS basis are what is being changed, which is precisely the mechanical layer a software tool can own. INFERENCE: the pain lands hardest on employers whose H-1B workers were priced at Level 1 or Level 2, since an upward re-tiering can force a wage increase, a re-filed LCA, or an amended petition.
Converging signals
Three things meet at one point: (1) a federal rule changing a published, deterministic formula; (2) a defined class of filers — every US employer sponsoring a foreign worker, plus the immigration firms and mobility teams that prepare their filings; and (3) government portals (INFERENCE: DOL FLAG and the OFLC Online Wage Library — the source text names the programs, the forms' underlying OEWS survey, and the four-tier structure, but does not name the portal) through which ETA Forms 9141, 9035, and 9089 must pass carrying a prevailing wage that survives the new computation. The second FORCED BUYER item in the input (the FLSA exemption technical amendment implementing court judgments vacating the 2024 overtime rule) is retrieval noise at cosine 0.728 — it is about FLSA white-collar exemptions, not immigration wages, and it does not support this opportunity. I am not citing it as demand evidence.
Customer pain
HYPOTHESIS (no complaint threads or job postings were supplied in the input for this specific rule): an immigration paralegal today prices a case by opening the OFLC Online Wage Library, entering a SOC code and area, reading four numbers, and matching the position's requirements to a tier via the DOL wage-level worksheet. Under a changed formula, every case in a firm's book of business has an unknown new answer, and the firm has no way to see its aggregate exposure without re-running each case by hand. That is a recomputation problem across hundreds or thousands of rows — exactly what software is for and exactly what a person doing it manually is slow and error-prone at.
Who pays
Immigration law firms (the highest-volume, most reachable buyer — they already buy per-seat case-management software), corporate in-house mobility teams at mid-size tech/staffing/healthcare employers, and immigration-focused consultants. NOT the government, and NOT a procurement office — this is bought on a credit card by a practice manager, which keeps it out of enterprise sales. INFERENCE: staffing firms and IT consultancies that sponsor at volume are the most acutely exposed, since their margins are set by the delta between the bill rate and the required wage.
Solved today
FACT (from source, by implication): the wage figure comes from the BLS OEWS survey exposed through DOL's wage library. INFERENCE, from general knowledge and not from the provided text: firms use the free OFLC Online Wage Library one lookup at a time, plus internal spreadsheets, plus the case-management suites (INSZoom/Mitratech, LawLogix/Equifax, Docketwise, Tracker Corp, Envoy Global) that already generate and file ETA forms. Some large firms have built internal wage calculators.
Why current solutions are bad
The free wage library answers one query at a time and only under the current formula. It cannot answer the question the rule actually creates, which is portfolio-shaped: 'across my 340 active LCAs, which ones break, by how much, and what does that cost me.' The incumbent case-management suites are slow to ship regulatory changes and price at the whole-suite level, so a firm cannot buy just the exposure analysis. HYPOTHESIS — I have not verified any incumbent's shipping cadence, and this is the crux assumption the whole business rests on.
Proposed product
A dual-formula prevailing-wage engine. Core: ingest the OEWS microdata and implement both the current four-tier computation and the NPRM's proposed computation, so any SOC code plus area of intended employment returns eight numbers (four tiers, two rules) plus the delta. Around that core, the actual product: an exposure scanner. An employer or firm uploads its case list (or the tool pulls the employer's own filings from DOL's public LCA disclosure files), and the tool returns a ranked report of which positions re-tier upward, the aggregate annual wage liability created, and which cases need an amended LCA or a re-run PWD request. Second phase: a form-prep layer that pre-populates ETA 9141/9035 with the surviving wage and walks the filer through the FLAG submission.
MVP version
Ship the dual-formula calculator plus the portfolio exposure report as a web app. Ingest: OEWS data files (public, static, downloadable) and the DOL OFLC quarterly LCA disclosure files (public, per-employer, per-case, includes SOC code, worksite, wage level, and offered wage). Implement both formulas. Accept a CSV upload or an employer-name lookup. Output a PDF: 'Under the proposed rule, N of your M positions re-tier; aggregate exposure $X.' Do NOT build FLAG submission in the MVP — the rule is not final and the portal integration is the expensive part. The MVP is a spreadsheet-killer with a PDF, and that is enough to charge for.
30-day build
Read the NPRM in full and implement both formulas exactly, with a test suite that reproduces the current OFLC wage library's published outputs for a few hundred sampled SOC-plus-area pairs. If the engine cannot reproduce today's official numbers, nothing else matters. Download and normalize the OEWS release and the last four quarters of OFLC LCA disclosure data. Confirm — do not assume — whether the NPRM's proposed computation is specified precisely enough in the text to implement deterministically; if it is not, the entire product thesis fails and you should stop here.
60-day build
Build the employer-lookup exposure report. Because the LCA disclosure data is public and names the employer, you can generate a personalized exposure report for any sponsoring employer in the country without them lifting a finger. Generate the top 500 most-exposed employers by aggregate delta. This is simultaneously the product and the lead list. Publish an aggregate analysis of the rule's national wage impact during the comment period and file a substantive comment; both are credibility and both are SEO against a query with no existing answer.
90-day revenue plan
Cold-send each of the 500 employers (and the 100 highest-volume immigration firms, also identifiable from the same disclosure data as case preparers) their own free one-page exposure summary with the detail behind a paywall. Sell a per-seat subscription to firms for unlimited recomputation and monitoring, and a per-report fee to employers who want a one-off. Revenue at 90 days is plausible but is NOT the base case — see kill arguments. INFERENCE: a realistic 90-day outcome is 10-30 paying seats and a few one-off reports, not a business.
Distribution path
This is the strongest part of the idea and it is where the founder's public-records edge lives. DOL publishes the LCA disclosure files quarterly with employer name, case count, SOC code, worksite, and wage level. That means the complete list of forced buyers is public, free, and already segmented by exposure. Cold outreach here is not spray-and-pray; it is 'here is your specific number.' Secondary: immigration-law CLE circuits, the AILA practitioner community, LinkedIn posts targeted at mobility managers during the comment period.
Pricing hypothesis
$199-399 per seat per month for firms (INFERENCE: sits below the whole-suite case-management spend and above a throwaway tool, and firms bill this to clients). $499-1,500 per one-off employer exposure report priced on case volume. Per-filing fee ($25-75) only in the phase-two form-prep layer, and only if the rule finalizes.
Technical difficulty
Low to moderate on the data engineering: OEWS and OFLC disclosure files are public, static, and well-structured. The real difficulty is legal-mechanical fidelity — implementing the tier computation exactly as the regulation specifies, and defending the number when a lawyer disagrees with it. One wrong wage level in a filing is a client's problem, not an inconvenience. Budget for a licensed immigration attorney as a paid reviewer of the engine's logic; this is exactly the kind of contractor spend the founder can now fund.
Legal / regulatory risk
Real and specific. Telling an employer what prevailing wage to certify on ETA Form 9141/9035/9089 sits close to the unauthorized practice of law. The safe posture is 'we compute the OEWS tier arithmetic; your attorney makes the determination,' with the tool positioned as sold TO attorneys rather than around them. Errors-and-omissions coverage and blunt disclaimers are mandatory, not optional. This is not the heavy_compliance flag (the founder need not become licensed to sell a calculator to lawyers), but it is a genuine constraint on how phase two can be marketed.
Platform dependency
None meaningful. The data is public federal data; the eventual submission target is a government portal with no owner who can deplatform you. Dependency is on DOL continuing to publish OEWS and OFLC disclosure files in a parseable form, which it has done for many years.
Founder fit
High on shape, and high on the two things that actually matter here: this is a public-records-to-product play, and the founder has already shipped a production tool that submits mandated filings into a federal portal (FMCSA Training Provider Registry) and charges per upload. PERM/LCA is structurally the same business one agency over. Fit is imperfect in one way that deserves saying plainly: the FMCSA buyer was a training provider with no counsel; this buyer is a law firm, and lawyers are a skeptical, credentialed, relationship-mediated market that will interrogate your numbers. The founder sells through demonstrated value, and a personalized exposure report IS demonstrated value — but the sales motion is harder than per-upload self-serve.
Breakout potential
Moderate. The engine generalizes: the same dual-formula-plus-portfolio-exposure pattern applies to any wage-determination regime (Davis-Bacon prevailing wages for federally funded construction, Service Contract Act wage determinations, H-2A adverse effect wage rates), each of which is a separate forced-filer class with public data and no good tooling. The wage engine is the asset; immigration is the first market, not the last. That optionality is the real reason to build this.
Final recommendation
CONDITIONAL BUILD — build the asset, do not yet build the business. This is a genuine public-money/forced-filer shape with a public, complete, pre-segmented buyer list, which is rare and valuable. But it is gated on an NPRM that may be vacated exactly as the input's own second document shows a prior DOL wage rule was vacated, and the calculator at its center is copyable in a week. The correct move is to spend 30 days building the dual-formula engine and the exposure report over public OEWS and OFLC disclosure data, publish the national exposure analysis during the comment period to own the search results and the practitioner mindshare, and sell one-off reports to the most-exposed employers immediately — while treating per-seat subscriptions and the FLAG filing layer as contingent on the rule reaching final. If the rule dies, the wage engine still ports directly to Davis-Bacon and Service Contract Act wage determinations, which are permanent, federally funded, and have their own forced-filer classes. That portability is what makes the 30 days worth spending; without it, this would be a bet on a proposed rule and I would say no.
Next action
Read the NPRM in full at the Federal Register URL and answer one question before writing any other code: does the proposed rule specify the new tier computation precisely enough to implement deterministically from the text, and what is the comment-close date? If the formula is under-specified or offered as multiple alternatives for comment, there is nothing to build yet — pivot the same engine to Davis-Bacon wage determinations instead.