ClearStamp — input-VAT recovery sentry for Nigerian buyers

1. One-liner

Watches every invoice a Nigerian company receives, flags the ones that will cost it input VAT, and chases the vendor to fix them before filing.

2. Trend signal — why now?

Nigeria’s Merchant-Buyer Solution (MBS / FIRSMBS) e-invoicing mandate is mid-rollout right now. Large taxpayers (₦5B+ turnover) hit hard enforcement on 1 July 2026; medium taxpayers (₦1B–5B) begin the same month with enforcement from January 2027; small businesses follow in 2028. Every valid invoice must be pre-cleared through MBS and carry a valid Invoice Reference Number (IRN) + QR stamp — and here’s the load-bearing fact: an invoice without a valid IRN is invalid for tax purposes, and the buyer loses the right to claim input VAT on that transaction entirely.

That turns your suppliers’ compliance into your cash-flow problem. As one Nigerian columnist put it plainly: “e-invoicing: manufacturer’s VAT now depends on supplier compliance.” A large taxpayer already compliant in Phase 1 will spend mid-2026 receiving non-compliant invoices from medium/small suppliers who onboarded late — and eating irrecoverable VAT on every one. Only ~1,000 businesses had reportedly completed onboarding by early 2026 against many thousands required. Penalties for the issuer are ₦200K per invalid invoice + ₦10K/day + a 100% surcharge on unreported transactions — which is exactly why buyers will refuse non-compliant invoices, and exactly why the buyer needs a system to catch them.

Accredited Access Point Providers (Interswitch, Duplo (YC W22, $5.6M raised), Beelpoint) are racing to sell the issuer side — “generate your invoice, route it to FIRS, settle payment.” Nobody is defending the buyer’s inbound side. That’s the gap.

Provenance:

• Signal 1 (Demand): Buyers lose input-VAT credit when suppliers send non-IRN invoices; “manufacturer’s VAT now depends on supplier compliance” — https://guardian.ng/opinion/columnists/e-invoicing-manufacturers-vat-now-depends-on-supplier-compliance/ + https://tryduplo.com/blog/nrs-e-invoicing-compliance-risks-nigeria — 2026-06/07
• Signal 2 (Feasibility): MBS now stamps every valid invoice with a machine-checkable IRN + QR; cheap multimodal AI reads any received PDF/photo/scan and verifies the stamp against the buyer’s own PO/GRN — no FIRS accreditation required for the buyer-side reader — https://einvoice.firs.gov.ng/ + https://www.cleartax.com/ng/e-invoicing-nigeria — 2026
• Signal 3 (Economic): ₦3.4T of VAT at stake; APPs funded and moving (Duplo dual-license, Interswitch accredited); ₦200K/invoice + 100% surcharge penalties quantify the spend — https://businessday.ng/opinion/article/the-n3-4-trillion-question-why-e-invoicing-is-bigger-than-compliance/ + https://techcabal.com/2026/03/16/duplo-secures-nrs-licence/ — 2026-03 Category: Regulatory arbitrage

3. The opportunity

Every accredited APP is selling the outbound story: help you issue compliant invoices and route them to FIRS. That’s the seller’s problem. But VAT is a two-sided ledger — your input-VAT recovery depends entirely on the invoices you receive, and those come from hundreds of vendors you don’t control, many of them smaller businesses onboarding late or badly.

Nobody owns the buyer’s inbound defense. The finance team’s current workaround is a junior accountant eyeballing PDFs, or discovering at filing time that ₦4M of input VAT is unclaimable because 60 supplier invoices had no IRN. ClearStamp is the sentry that sits on the buyer’s AP inbox: it reads every incoming invoice, verifies the IRN/QR stamp is real and matches the transaction, catches the missing/fake/mismatched ones the day they arrive, and auto-nudges the vendor to reissue a valid one before the buyer pays or files.

Incumbents (APPs) won’t build this — it’s orthogonal to their payments/issuer product and doesn’t need their accreditation, so it’s a low-status feature for them and a whole product for us. Generic AP tools (Zahara, QuickBooks) don’t know what an IRN is.

4. Target market

• Primary customer: Finance manager / tax lead at a Nigerian VAT-registered buyer with ₦1B–₦50B turnover — manufacturers, distributors, hospitality groups, construction/EPC firms — that receives 100–2,000+ supplier invoices a month from a long tail of vendors.
• Why they buy: “I only find out an invoice has no IRN when I’m doing my VAT return and it’s too late — I’ve already paid the vendor and now I’m eating the VAT.” Recoverable input VAT lost to one late supplier is real, monthly, quantifiable cash.
• Rough TAM reasoning: Large taxpayers number in the low thousands; medium taxpayers (₦1B–5B) tens of thousands. Even 3,000–8,000 reachable mid-to-large buyers at ₦150K–₦400K/mo is a comfortable sub-$5M ARR pond.
• Why now for them: July 2026 enforcement means their supplier base is simultaneously onboarding — the exact window where non-compliant inbound invoices spike. This pain has a start date, and it’s this month.

5. Product sketch (MVP)

• Inbound invoice capture: forward your AP inbox / drop a folder — ClearStamp ingests every supplier invoice (PDF, scan, photo, email body).
• IRN + QR verification: reads the stamp, confirms it’s a real FIRS-issued IRN and not a copy-paste fake or a Phase-1-style unstamped invoice.
• Three-way match: ties each invoice to your PO and goods-received note; flags mismatched amount, TIN, or line items that would fail clearance.
• Traffic-light AP queue: green = safe to pay & claim, amber = fixable, red = will cost you input VAT if you pay it as-is.
• Auto vendor chase: one-click (or automatic) WhatsApp/email to the supplier: “This invoice has no valid IRN; please reissue a cleared one or we can’t pay — here’s how.” Tracks who’s fixed it.
• Vendor readiness board: which of your suppliers are MBS-ready, which aren’t, ranked by the ₦ of VAT you have at risk with each.
• Monthly recovery report: input VAT protected, VAT at risk, worst offenders — the number the CFO screenshots.

6. AI angle — what’s load-bearing

AI does the actual work: reading thousands of heterogeneous supplier invoices (no standard format — Word templates, scanned paper, phone photos, PDFs in ten layouts), extracting IRN/QR/TIN/amounts, matching them to POs/GRNs across naming and rounding noise, and drafting the vendor-chase message in the right tone. Remove the AI and you’re back to a junior accountant eyeballing PDFs one by one — which is the exact status quo this replaces. The IRN validity check is a rules layer; the extraction and matching at document scale is the AI, and it’s the whole product.

7. Localization angle

This is the localization play — it’s Nigeria-specific regulatory arbitrage. The MBS/IRN scheme, ₦200K penalty structure, the July-2026 supplier-readiness cliff, and vendor-chase via WhatsApp (Nigeria’s default B2B channel) are all local. Pricing in naira (₦150K–₦400K/mo) works where a $500/mo global AP tool would be laughed out of the room. The same template ports to any post-2024 CTC (continuous transaction control) mandate — Malaysia MyInvois, Philippines EIS, Saudi ZATCA all create the identical buyer-side input-credit exposure — so Nigeria is the beachhead, not the ceiling.

8. Business model — path to $1M–$5M ARR

• Pricing: ₦150K/mo (≈$95) for buyers under ~300 inbound invoices/mo; ₦400K/mo (≈$255) for high-volume (2,000+ invoices, multi-entity). Annual VAT-at-risk report justifies it in one line.
• ACV: ~₦3.6M (≈$2,300) blended.
• Rough math to $1M ARR: ~440 buyers × ₦3.6M/yr ≈ ₦1.6B ≈ $1M. Reachable inside the large + medium taxpayer base.
• Rough math to $5M ARR: ~2,200 buyers, or hold buyer count and add a per-vendor-onboarding assist fee + expand to one more CTC country (Malaysia/Philippines). Realistic in 24–30 months given the enforcement tailwind.
• Expansion path: multi-entity seats → vendor-onboarding concierge (help your suppliers get MBS-ready, charged per vendor) → outbound issuance via an APP partner (revenue share, no accreditation of our own) → adjacent CTC geographies.

9. Go-to-market wedge — first 100 customers

• Big-4 & mid-tier tax advisors as channel: Nigerian tax practices (Andersen, Ascension, PwC/KPMG mid-market teams) are fielding panicked “am I going to lose my VAT?” calls all quarter. Offer them a co-branded ClearStamp report to hand clients; they bring us the buyers. Sign 5 firms → dozens of warm intros each.
• Scrape the enforcement cliff: the medium-taxpayer list (₦1B–5B) is definable via public filings, MAN (Manufacturers Association of Nigeria) directories, and sector associations. Cold-email the finance lead a free “VAT-at-risk audit” of one month of their AP inbox — show them the ₦ they’re about to lose. That number closes.
• Ride the APP gap: Duplo/Interswitch onboard the issuer; publicly partner or content-market as “the inbound half they don’t do.” LinkedIn + Nairametrics/TechCabal op-eds aimed at CFOs during the exact July–December 2026 enforcement window.
• MAN / sector WhatsApp groups: Nigerian manufacturing and distribution run on WhatsApp broadcast groups; a single well-placed “here’s how much input VAT you’re about to lose” explainer travels.

10. Build complexity — justification

Medium. Document ingestion + multimodal extraction + IRN/QR validation + PO/GRN matching + a vendor-chase workflow is standard AI-app engineering on off-the-shelf models and web stack — no custom ML, no FIRS accreditation, no capital-heavy integrations. The gnarly parts are (a) robust extraction across messy real-world Nigerian invoice formats and (b) confirming IRN validity (public QR/stamp verification, not privileged FIRS API access). A focused pair ships a credible v1 in ~10–14 weeks.

11. Gating checklist

Gate	Pass?	Note
Legal in target market	✅	Reading a buyer’s own received invoices; no accreditation needed to verify a stamp.
Ethical — no harm / dark patterns	✅	Helps buyers claim legitimate VAT and pushes suppliers toward compliance. Net-positive.
Market exists (evidence above)	✅	Active mandate, quantified penalties, funded incumbents on the adjacent side.
1–5 person team can build this	✅	Medium complexity, off-the-shelf AI.
Launchable with <$50K / ₹40L	✅	No hardware, no accreditation, no licensing capex.

12. Feasibility score

Axis	Weight	Score	Notes
Problem intensity	20	16/20	Real, recurring, quantified in ₦ — but it’s a “lose money you didn’t have to” pain, slightly softer than “business stops today.”
Demand evidence	15	12/15	Strong: active mandate, columns explicitly naming the buyer-side VAT loss, funded APPs. Docked because the buyer-side product has no direct proven spend yet — the APPs prove issuer-side demand.
Build feasibility	15	12/15	Off-the-shelf AI; risk is extraction robustness on chaotic invoice formats.
Distribution clarity	15	11/15	Tax-advisor channel + free-audit cold play is concrete; conversion unproven.
Revenue mechanics	15	12/15	₦ pricing benchmarked, ACV realistic, path clear; churn after enforcement panic settles is the risk.
Time to first revenue	10	7/10	Pain is live now; a free-audit-to-paid funnel can close in 6–8 weeks.
Defensibility	10	5/10	Execution + workflow lock-in + vendor-readiness data that compounds — but an APP could bolt this on. Speed and CFO trust are the moat.
Total	100	75/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy · domain-expertise-required — needs someone who can build robust document AI and someone who genuinely understands Nigerian VAT mechanics and the MBS scheme (or a tax advisor co-founder).

Key assumptions to validate (3–5)

1. Assumption: Buyers will pay ₦150K–₦400K/mo to protect input VAT they’d otherwise lose. How to test: Run 20 free VAT-at-risk audits on real AP inboxes; count how many convert to paid after seeing their ₦ figure.
2. Assumption: IRN/QR validity can be verified buyer-side without privileged FIRS API access. How to test: Confirm the public QR/stamp verification path on 50 real cleared invoices before writing a line of product code.
3. Assumption: Extraction is robust enough across messy Nigerian invoice formats to be trusted for a VAT decision. How to test: Collect 500 real received invoices from 3 pilot buyers; measure extraction + IRN-detection accuracy.
4. Assumption: Tax advisors will channel-partner rather than build it themselves. How to test: Pitch 5 mid-tier firms; see if they’ll co-brand a report.

Risk flags

1. Platform/regulatory dependency: FIRS could publish an official buyer-side validation portal that commoditizes the check — collapsing the “verify the stamp” value. Mitigate by owning the workflow (matching + vendor chase + recovery reporting), not just the check.
2. Market timing: Once every supplier is onboarded (2028+), inbound non-compliance decays — the acute pain has a shelf life. Mitigate by pivoting the same rails to ongoing three-way-match/AP-audit value and to newer-mandate geographies.
3. Incumbent bolt-on: An APP (Duplo/Interswitch) adds inbound verification as a feature. Mitigate with CFO trust, the vendor-readiness dataset, and being the neutral buyer-side tool that isn’t tied to one issuer.

14. Structured verdict

Score:                  75/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Technical founder (document AI) + Nigerian VAT/tax domain co-founder or advisor
Time to revenue:        6–10 weeks (free-audit → paid funnel)
Capital to launch:      ₦4–7M / $3–5K
Top 3 assumptions to validate first:
  1. Buyers convert to paid after seeing their ₦ VAT-at-risk number — 20 free audits
  2. IRN/QR is verifiable buyer-side without privileged FIRS access — test on 50 cleared invoices
  3. Extraction accuracy on messy real invoices is high enough to trust — 500-invoice benchmark
Kill criteria:
  - Abandon if <15% of 20 free-audited buyers convert to paid within 30 days
  - Abandon if IRN validity cannot be confirmed without becoming an accredited APP
  - Abandon if FIRS ships a free official buyer-side validation portal that also handles vendor chase

15. Next step — 1-week validation sprint

• Day 1–2: Confirm the buyer-side IRN/QR verification path works without accreditation — grab 50 real cleared Nigerian invoices, verify their stamps end-to-end. If this fails, the idea is dead; find out first.
• Day 3–4: Line up 5 mid-to-large Nigerian buyers (via a tax-advisor intro or LinkedIn) and run a manual “VAT-at-risk audit” on one month of their received invoices — by hand if needed. Put a ₦ figure on their at-risk input VAT.
• Day 5: Decide go / no-go on a falsifiable result: do ≥2 of 5 audited buyers say “yes, I’d pay ₦150K+/mo to never see this number surprise me again”? If yes, build. If no, the pain isn’t sharp enough to sell.