GO
Overall Score
CreditKeeper — hybrid-regime navigator for Brazilian Simples providers
1. One-liner
Shows a Brazilian Simples provider which business clients the tax reform costs them — and whether to switch by September.
2. Trend signal — why now?
Brazil’s tax reform (Reforma Tributária, LC 214/2025) starts biting in 2026. Two consumption taxes — IBS and CBS — replace PIS, COFINS, ICMS and ISS in a phased rollout to 2033. The detail that matters for small providers is buried in the credit rules: a Simples Nacional (or MEI) provider who keeps paying the new taxes inside the unified DAS bill passes only a sliver of IBS/CBS credit to a business (PJ) buyer, while a regular-regime supplier passes the full ~28%. So a corporate client that can claim credit has a hard, purely-tax reason to drop the Simples vendor for one that isn’t.
The reform gives Simples firms an escape hatch: opt for the hybrid regime (“por fora”) — compute IBS/CBS in the regular regime so clients get full credit, while keeping IRPJ, CSLL, CPP inside the DAS. The catch is the deadline. This year the Simples option for 2027 moves from January to September 1–30, 2026 — one window, one decision, and it locks January–June 2027. Miss it or guess wrong and a B2B provider watches clients migrate without ever being told why.
Every tool that exists today is a seller-side price simulator (Conta Azul, Contmatic, Taxcel, simulareforma). None of them answer the question the provider actually loses sleep over: given MY client list, do I lose business if I stay simple, and is opting hybrid worth it?
Provenance:
- Signal 1 (demand): “Reforma Tributária acende alerta para MEIs e empresas do Simples que vendem para outras empresas” — small B2B providers risk losing corporate clients over credit rules — https://www.diarioro.com.br/2026/07/08/reforma-tributaria-acende-alerta-para-meis-e-empresas-do-simples-que-vendem-para-outras-empresas/ — 2026-07-08
- Signal 2 (feasibility/economic): Receita/CGSN sets the September 1–30, 2026 hybrid-regime decision window; being current on taxes is a precondition — https://www.contabeis.com.br/noticias/75553/simples-nacional-deve-decidir-sobre-regime-hibrido-ate-setembro-de-2026/ — 2026 (accessed 2026-07-21)
- Signal 3 (market size): MEI = 16,291,125 active registrations (63.3% of formal businesses); Simples Nacional = 7,348,088; together >90% of Brazilian companies — https://www.meirafernandes.com.br/noticias/mei-e-simples-nacional-concentram-919-das-empresas-no-brasil-enquanto-lucro-real-e-presumido-apenas-68/ — 2026 Category: Regulatory arbitrage
3. The opportunity
The incumbents are tax calculators. Feed them a sale price and a rate and they spit out a new price to protect your margin. That’s a pricing question. The unaddressed question is a relationship question: which of my clients have a tax reason to leave, and does opting hybrid keep them?
That answer can’t be pre-computed. It depends on the provider’s specific mix — how much revenue comes from PJ buyers who can use credit vs. PF/consumer buyers who can’t, sector, margin, and how much extra tax/complexity the hybrid regime adds for this provider. Working that out today means a rushed conversation with an overloaded contador who has hundreds of these to run before September and defaults to generic advice.
CreditKeeper reads the provider’s actual invoices and client list, classifies each client by whether they benefit from full credit, and returns a plain-language verdict: stay simple (and here’s the revenue at risk), or go hybrid (and here’s the net trade-off). Then it drafts the client-facing note explaining the change. It’s a decision-support tool aimed squarely at the September window — not another rate calculator.
4. Target market
- Primary customer: Brazilian Simples Nacional and higher-revenue MEI service providers who sell mostly to businesses — IT/dev freelancers and small agencies, marketing/design shops, consultancies, cleaning & facilities, logistics/transport, technical maintenance. Revenue roughly R$100K–R$4.8M/yr, 1–20 people, an external contador.
- Why they buy: “I keep hearing I’ll lose corporate clients under the reform and I don’t understand why or whether I should change regime. My accountant says ‘depends’. September is coming.” It’s a money-on-the-table decision with a deadline and no clear tool.
- Rough TAM reasoning: 7.35M Simples firms + 16.3M MEIs. Service providers are a large share; the B2B-selling subset is the target. Even 300K–800K reachable B2B service providers is a deep pool for a bootstrapped tool priced at R$39–149.
- Why now for them: The September 1–30, 2026 hybrid-regime option is a one-shot decision that locks H1 2027. The urgency is real and dated, not evergreen.
5. Product sketch (MVP)
- Import notas fiscais (NFS-e/NF-e XML or a CSV export) and a client list; auto-classify each client as credit-benefiting PJ vs. non-benefiting PF/consumer.
- “Revenue at risk” view: how much of your billing comes from clients who now have a tax reason to prefer a full-credit supplier.
- Stay-simple vs. go-hybrid comparison for your numbers, in reais, with the net trade-off spelled out (client-retention upside vs. added tax/complexity).
- Plain-Portuguese verdict and a checklist for the September portal step if hybrid wins.
- Auto-drafted client note (“why our nota will now show IBS/CBS you can credit”) to send to the clients most at risk.
- Contador-share mode: export the analysis as a one-page PDF the accountant can sign off, instead of building it from scratch.
6. AI angle — what’s load-bearing
Two jobs. First, classification from messy real data: reading heterogeneous nota fiscal exports and client records and inferring, per client, whether they’re a credit-using PJ — names, CNPJ/CPF patterns, product/service codes, historic behavior. Second, explanation: turning a per-client tax verdict into language a non-accountant provider and their client both understand, and drafting the client note. Remove the AI and you’re back to a static rate calculator that can’t reason over an individual’s client mix — which is exactly the commodity that already exists and doesn’t solve the problem. The reasoning-over-your-own-data step is what only became cheap enough to productize in the last year.
7. Localization angle (if any)
This is intrinsically Brazil-only and that’s the moat, not a limitation. It encodes LC 214/2025’s credit mechanics, the Simples vs. hybrid distinction, the September 2026 window, nota fiscal (NFS-e/NF-e) formats, and Portuguese client-facing copy. Pricing in reais (R$39–149/mo) works where a US$49 tool wouldn’t. Distribution runs through contadores and MEI/Simples communities. A generic global tax tool cannot touch this.
8. Business model — path to $1M–$5M ARR
- Pricing: R$49/mo self-serve for a single provider; R$149/mo “contador” tier that runs many clients through one dashboard and exports signed PDFs. A one-time R$99 “September decision” report for providers who only want the verdict once.
- ACV: Blended ~R$1,000/yr (mix of monthly subs and one-time reports; contador tier pulls the average up).
- Rough math to $1M ARR (~US$180K–200K): ≈R$5.4M ARR. ~4,500 providers at R$1,000/yr blended, OR ~600 contadores at the R$149 tier managing books × 12. Against 300K+ reachable B2B service providers and tens of thousands of contadores, reachable.
- Rough math to $5M ARR: Requires the tool to persist past the September event — recurring re-evaluation each transition year (2027, 2029), plus expansion into ongoing correct-credit invoicing checks for hybrid-regime providers, and a contador-channel that resells it. Needs the one-shot decision to become a subscription habit.
- Expansion path: Event tool → annual regime re-evaluation → “is my nota passing the right credit?” monitoring as the reform phases up (CBS 0.9% in 2026 → 8.8% in 2027) → payroll/PJ-contract advisory adjacent.
9. Go-to-market wedge — first 100 customers
- Contador channel (primary): Accountants are drowning in the September decision for every Simples client. Recruit 20–40 via CRC state councils and accounting Facebook/WhatsApp groups; give them the R$149 dashboard free through September in exchange for running their B2B clients through it. Each contador carries dozens of target providers.
- Complaint-adjacent content in Portuguese: The “vou perder clientes” fear is already circulating (Diário RO, CRCSP, SIMPI, YouTube explainers like “A cilada da Reforma Tributária para quem presta SERVIÇOS”). Publish a free “will you lose B2B clients?” mini-check that captures the provider’s rough mix and produces a teaser verdict → paid full report.
- Vertical communities: Dev/agency and cleaning/logistics provider groups on WhatsApp/Telegram and sector associations (e.g. SIMPI’s micro-industry base). Post the free check where the anxiety already lives.
- Timing sprint: July–September 2026 is the acquisition window. The message writes itself: “Decide before September 30, or your clients decide for you.”
10. Build complexity — justification
Low. The core is document parsing (nota fiscal XML/CSV), a rules layer encoding the Simples-vs-hybrid credit logic from LC 214/2025, an LLM for per-client classification and Portuguese explanation, and a simple web app. Everything is off-the-shelf; no custom models, no bank/government integrations required for v1 (manual/CSV import is fine). A solo builder with a Brazilian tax advisor could ship v1 in 6–8 weeks — the constraint is domain accuracy, not engineering.
11. Gating checklist
| Gate | Pass? | Note |
|---|---|---|
| Legal in target market | ✅ | Decision-support/advisory tool; disclaims that final filing is the contador’s call. |
| Ethical — no harm / dark patterns | ✅ | Helps small providers see a real risk they’re otherwise blind to. |
| Market exists (evidence above) | ✅ | 23.6M MEI+Simples firms; documented “losing clients” alarm; dated Sept 2026 decision. |
| 1–5 person team can build this | ✅ | Solo builder + tax advisor. |
| Launchable with <$50K / ₹40L | ✅ | Web app + API costs; well under budget. |
All five pass.
12. Feasibility score
| Axis | Weight | Score | Notes |
|---|---|---|---|
| Problem intensity | 20 | 16/20 | Dated, money-losing decision with a hard deadline; providers are anxious but many don’t yet grasp the credit mechanics — so pain is real but partly latent, which is why it’s not a full 18. |
| Demand evidence | 15 | 11/15 | Strong structural + regulatory signal and clear press alarm, but verbatim owner quotes were thin (mostly institutional voices: SIMPI, CRCSP, SEBRAE). Downgraded honestly for that. |
| Build feasibility | 15 | 13/15 | Off-the-shelf stack, 6–8 weeks; only risk is encoding the tax logic correctly. |
| Distribution clarity | 15 | 12/15 | Contador channel is concrete and leveraged; conversion via accountants is plausible but unproven. |
| Revenue mechanics | 15 | 11/15 | Pricing fits Brazilian wallets; risk is that it reads as a one-time purchase, not a subscription. |
| Time to first revenue | 10 | 7/10 | The Sept window pulls revenue forward fast, but the same window is a cliff — see risk flags. |
| Defensibility | 10 | 3/10 | Copyable; the only moat is speed to the September window and the contador relationships. Incumbent simulators could add a client-mix view. |
| Total | 100 | 73/100 |
13. Qualitative modifiers
Founder-fit tags
domain-expertise-required (must encode LC 214/2025 credit mechanics correctly — a tax advisor co-founder or close advisor is non-negotiable) · technical-heavy (document parsing + LLM classification).
Key assumptions to validate (3–5)
- Assumption: B2B Simples service providers will pay to decide, not just to file. How to test: offer the R$99 one-shot report to 50 providers sourced from accounting groups; measure paid conversion before building the full app.
- Assumption: Contadores will adopt and push the tool rather than see it as encroachment. How to test: pitch 15–20 accountants; count how many will run even one client through a prototype.
- Assumption: Real nota fiscal / client exports can be classified into credit-benefiting vs. not with usable accuracy. How to test: run 10 real providers’ exports by hand vs. the model; measure agreement.
- Assumption: The September deadline creates urgency-to-pay, not urgency-to-ask-my-accountant-for-free. How to test: A/B the free mini-check’s paywall conversion in August.
Risk flags
- Market timing (event cliff): The sharpest demand is a single September 2026 window. After it closes, the pitch weakens until the next transition year. If the recurring/monitoring product isn’t ready, this is a spike, not a business.
- Platform/regulatory dependency: The reform is phased and the rules are still being clarified by Receita/CGSN through 2026. A late regulatory tweak to Simples credit treatment could reshape the core logic — must track official guidance continuously.
- Channel-conflict / substitution: Contadores could decide this is their job and do it in a spreadsheet, and incumbent simulators (Conta Azul et al.) could bolt on a client-mix view. Defensibility is thin; win on focus and speed or don’t play.
14. Structured verdict
Score: 73/100
Verdict: GO
Confidence: Medium
Best-fit builder: Technical solo founder + Brazilian tax advisor (contador/tributarista)
Time to revenue: 4–8 weeks (pre-sell R$99 report before full build)
Capital to launch: R$20K–40K (US$4–7K) — mostly the founder's time + API costs
Top 3 assumptions to validate first:
1. Providers pay to DECIDE, not just to file — sell 50 one-shot reports before building
2. Contadores adopt rather than resist — pitch 15–20, count prototype runs
3. Client-mix classification from real notas is accurate enough — hand-check 10 providers
Kill criteria:
- Abandon if <5% of 50 targeted B2B providers pay R$99 for the one-shot report by end of August 2026
- Abandon if <3 of 20 pitched contadores will run a single client through the prototype
- Abandon if Receita guidance removes the Simples-vs-hybrid credit gap (the entire premise)
15. Next step — 1-week validation sprint
- Day 1–2: Build the free “will you lose B2B clients?” mini-check (5 questions: revenue, % from PJ clients, sector, margin, current regime) that outputs a teaser verdict. No backend tax engine yet — hardcode the logic with the advisor.
- Day 3–4: Drop it into 5–8 accounting and provider WhatsApp/Facebook groups and DM 30 B2B providers who’ve posted about the reform. Offer the R$99 full one-shot report behind the teaser.
- Day 5: Decide go/no-go on a falsifiable number — ≥5 paid R$99 reports (or ≥5 firm pre-commitments) AND ≥3 contadores willing to run a client through a prototype. Below that, the September spike isn’t monetizable and it’s a PASS.
Interested in a detailed proposal?
Get a deep-dive with market research, competitive analysis, and implementation roadmap.
Contact usinfo@startupbasket.ai