What changed
FACT (from input hypothesis, consistent with public IRS practice): the IRS publishes the Exempt Organizations Business Master File with addresses, the auto-revocation list, and 990/990-N filing indexes, making an at-risk cohort computable from public data. HYPOTHESIS: Microsoft's nonprofit grant withdrawal (referenced as signal 1729, not included in this input) is forcing small nonprofits to re-examine their back office right now, creating an outreach window.
Why now
Three consecutive missed 990-series filings trigger automatic revocation of exempt status — a hard statutory deadline (IRC 6033(j)), so every two-strike org has exactly one filing season left. The claimed M365 wedge is an inference with no signal text provided here, so treat the timing hook as unproven; the revocation clock alone is real and permanent.
Converging signals
Input provided ZERO signals and ZERO demand_evidence items. The convergence is a pattern-transfer hypothesis (public registry + recurring filing duty + market-exclusion penalty), not an observed signal cluster. The system's own lesson (confidence 0.85) that it is capability-rich but demand-blind applies: absence of demand evidence here may be an ingestion gap, but per instructions it is scored as absent, not assumed.
Customer pain
HYPOTHESIS: a volunteer treasurer at a small nonprofit that has missed two consecutive 990-N filings faces automatic revocation after year three — donors lose deductibility, fiscal sponsors and grant platforms (Benevity, Fidelity Charitable) drop the org, and reinstatement requires Form 1023 refiling plus fees, a far more painful process than the 10-minute e-postcard that would have prevented it. The pain is severe but LATENT: by definition this cohort has ignored the duty twice already.
Who pays
Board members/treasurers of small US nonprofits (mostly under $50k gross receipts, hence 990-N filers). Secondary payers: fiscal sponsors, community foundations, and accountants managing multiple small orgs who need portfolio monitoring. No demand_evidence was provided proving willingness to pay; incumbent paid-filing services existing at all is the main indirect evidence.
Solved today
FACT: the IRS itself offers 990-N e-filing free, online, in minutes. Paid alternatives exist: Tax990/TaxBandits (~$20/filing), File990, Aplos. Status monitoring is DIY via IRS Tax Exempt Organization Search. Nobody visibly does proactive outbound to the computed two-strike list.
Why current solutions are bad
The free IRS route requires the org to remember and act — the two-strike cohort demonstrably does not. Incumbents are inbound-only: they serve orgs that already know they must file. Nothing watches an org's status and intervenes before strike three. The wedge is targeting + monitoring, not the filing mechanics.
Proposed product
Revocation Shield: (1) nightly diff of IRS BMF, 990/990-N filing indexes, and the auto-revocation list; (2) scored at-risk registry (one-strike, two-strike, revoked); (3) subscription that monitors the org's EIN, alerts on deadlines, and files the 990-N (via an existing e-file API such as TaxBandits rather than pursuing direct IRS MeF authorization); (4) reinstatement-guidance upsell for already-revoked orgs; (5) portfolio dashboard tier for accountants and fiscal sponsors.
MVP version
A pipeline that ingests public IRS data and outputs the two-strike list with addresses, plus a one-page 'check your revocation risk by EIN' site and a concierge filing service (manually assisted first 20 customers, TaxBandits API after). No IRS e-file authorization needed for MVP.
30-day build
Build the cohort pipeline; validate the testable prediction (is the two-strike cohort >=50k orgs, and what fraction show recent website/state-registration activity vs defunct); stand up EIN-lookup landing page; send 500-piece direct-mail/email test to the highest-liveness two-strike orgs.
60-day build
If conversion >=0.5-1%, wire TaxBandits (or equivalent) API for automated 990-N submission; add status-change monitoring and deadline alerts; begin outreach to accountants/fiscal sponsors with the portfolio view.
90-day revenue plan
Scale outbound against the full cohort in waves timed to each org's fiscal-year deadline; add $149 'rescue' one-time product for orgs at strike three or freshly revoked (reinstatement packet). Target: 150-300 subscribers at $99/yr plus rescue fees.
Distribution path
Precision outbound (postal mail + email) to addresses already in the public BMF — the target list IS the product's own dataset, which is the rare case where cold outreach is perfectly targeted. Secondary: SEO on 'nonprofit status revoked' / EIN-lookup tool as lead magnet; partnerships with state nonprofit associations.
Pricing hypothesis
$99/yr monitoring + filing (vs ~$20 one-off at incumbents — justified by monitoring, alerts, and done-for-you continuity); $149 one-time rescue/reinstatement packet; $499+/yr accountant/fiscal-sponsor portfolio tier.
Technical difficulty
Low-moderate and squarely in founder's lane: public-data ETL, diffing, scoring, outbound automation — the same public-records + government-filing shape as his shipped FMCSA ELDT product. Key simplification: ride an existing IRS-authorized e-file API instead of becoming a MeF transmitter. Main technical risk is data freshness — IRS publishes 990-N filing data with lag, so 'exactly two missed years' detection will be noisy at the margins.
Legal / regulatory risk
Moderate-low. 990-N is a simple informational e-postcard, not tax preparation with liability exposure; using an authorized e-file provider's API offloads transmitter compliance. Must avoid implying IRS affiliation in outbound (FTC/impersonation concerns) — mailers to compliance-deadline lists are a scrutinized category, so copy must be clearly commercial.
Platform dependency
Depends on continued IRS publication of BMF/filing indexes (long-standing, statutorily grounded — low risk) and on a third-party e-file API (substitutable). No app-store or social-platform exposure.
Founder fit
High. Applying the system's government-portal lesson (confidence 0.80): this matches the proven ELDT pattern — federal duty, identifiable forced filers, submission automation, per-transaction/subscription monetization — and it is public-records + compliance-monitor territory he prefers. One structural difference vs ELDT keeps this from a 9: the government alternative here is FREE and easy, so he is not removing a painful portal, he is removing forgetting.
Breakout potential
Good expansion surface: state charity-registration renewals (41 states, genuinely painful, often paid), registered-agent-style compliance bundles, grant-eligibility monitoring for funders, and selling the at-risk dataset itself to accountants and nonprofit consultants.
Final recommendation
CONDITIONAL GO — this is a cheap, well-shaped test, not yet a validated business. Zero demand evidence was provided, so the entire case rests on the structural forced-buyer logic. Run the falsification plan from the hypothesis first: compute the cohort, measure liveness (websites, state registrations), and run the 500-org outbound test before writing product code beyond the pipeline. Budget: <$2k and ~3 weeks to a kill/scale decision.
Next action
Download IRS EO BMF + 990-N/990 e-file indexes + auto-revocation list; compute the two-consecutive-missed-years cohort and a liveness score; if >=20k live-looking orgs, send the 500-piece outbound test with an EIN risk-check landing page.