SB StartupBasket
All ideas
76 /100 GO Medium complexity

PixFiel — recurring-revenue medic for Brazil subscriptions

Diagnoses why every Pix Automático charge silently failed and runs the WhatsApp script that rescues the mandate before churn.

FinTech SaaS LATAM WhatsApp-firstPix-nativeCompliance-drivenSMBSolo-builderCreator
— views
Evaluation Scores
76/100

GO

Overall Score

16
Problem
12
Demand
11
Build
12
Distrib.
12
Revenue
8
Time
5
Defense

PixFiel

1. One-liner

Diagnoses why every Pix Automático charge silently failed and runs the WhatsApp script that rescues the mandate before churn.

2. Trend signal — why now?

Brazil’s Central Bank launched Pix Automático in June 2025 — the first real account-to-account recurring-payment rail in the country. Adoption is not gradual, it’s vertical: transaction volume grew 182% quarter-over-quarter between Q4 2025 and Q1 2026, and by mid-2026 Pix Automático already accounts for 19–28% of subscription payment volume at merchants that switched. It beats cards on approval rate by ~20 percentage points.

But the rail has a brand-new failure mode nobody has tooled for. The customer — not the merchant — sets a per-charge ceiling and a daily Pix Automático limit inside their own bank app, and can cancel or suspend the mandate silently at any time. A charge above the ceiling fails or flags. Insufficient funds fails. A hit daily limit fails. The merchant just sees “failed” in the PSP dashboard with no why-in-mandate-terms and no fix. The entire recovery ecosystem today (Hotmart’s Recuperador, Vindi/Mercado Pago dunning playbooks, WhatsApp recovery bots) was built for card failures — expired cards, declined authorizations — and is blind to mandate semantics.

Provenance:

3. The opportunity

Every recurring seller on Pix Automático loses a slice of MRR each cycle to failed charges. On cards, an “expired card” retry logic recovers most of it. On Pix Automático the failure reasons are different in kind — the customer’s mandate ceiling is too low for your R$97 plan, their daily limit is exhausted, or they killed the authorization in Nubank/Itaú without telling you. Generic card dunning can’t diagnose or fix any of these, because the fix isn’t “retry the same charge” — it’s “text the customer the exact steps to raise their ceiling / re-authorize the mandate in their bank app.”

The PSPs (Asaas, Vindi, Iugu, dLocal, EBANX) own the rail and won’t build this — for them a failed mandate is one status field in a dashboard, and they serve API-integrated merchants. The gap is the workflow layer for the seller who is not a developer: they’re on a no-code checkout (Hotmart, Kiwify, Guru, Eduzz) or a light PSP account, they see “falhou,” and they have no idea why or what to say. PixFiel is a focused AI-first tool that reads the mandate/charge event stream, classifies the mandate-specific failure reason, and drives the recovery conversation on WhatsApp — the channel every Brazilian actually answers.

4. Target market

  • Primary customer: Brazilian subscription sellers doing roughly R$20k–R$300k/mo in recurring revenue on Pix Automático — infoproduct/course creators on Hotmart/Kiwify/Guru, subscription-box operators, micro-SaaS, gyms/CrossFit boxes, clubs/associations, community/membership sellers. 1–3 people, no in-house dev.
  • Why they buy: “I switched to Pix Automático because approval is better than cartão, but now when a charge fails I just get ‘recusado’ and my faturamento drops — I don’t know if the client cancelled, ran out of limit, or set the ceiling too low, and I’m chasing people one-by-one on WhatsApp.” Recovered MRR is the pitch — it pays for itself.
  • Rough TAM reasoning: Kiwify alone reports 29k+ creators; Hotmart is larger; add Guru/Eduzz + non-infoproduct subscription SMBs (gyms, boxes, associations) and the pool of recurring sellers big enough to feel failed-mandate pain is comfortably in the low-to-mid hundreds of thousands. Even 0.5% penetration at R$150/mo is a real business.
  • Why now for them: They migrated to Pix Automático in the last 12 months. The failure volume only became material once the rail crossed ~20% of their TPV — which is exactly mid-2026.

5. Product sketch (MVP)

  • One-click connect to their checkout/PSP (Hotmart, Kiwify, Guru webhooks; Asaas/Vindi/Iugu APIs) — no code, paste a key.
  • Mandate health board: every active Pix Automático mandate with a status — healthy / ceiling-too-low / limit-risk / suspended / cancelled-in-bank / insufficient-funds-pattern.
  • Failure diagnosis: when a charge fails, PixFiel classifies why in mandate terms and tells the seller in plain Portuguese what actually happened.
  • WhatsApp recovery flows: auto-sends the exact fix script for each failure type (“seu limite do Pix Automático está em R$50, aumente para R$97 assim: …” with bank-specific steps for Nubank/Itaú/Bradesco/Caixa) and books the re-authorization.
  • Smart retry orchestration: fires the allowed 3-retries-in-7-days window at the right moments (after payday, after the customer confirms a top-up) instead of blindly.
  • Pre-emptive ceiling nudge: flags mandates whose authorized ceiling is below the next scheduled charge (e.g. an annual price bump) and fixes them before they fail.
  • Recovered-MRR dashboard: shows R$ recovered this month vs. the subscription fee — the retention hook.

6. AI angle — what’s load-bearing

Two places AI does real work. First, failure classification: the raw signal is a terse, inconsistent status code from whichever PSP/checkout the seller uses, plus the mandate’s history (past charge amounts, ceiling, limit hits, timing). Mapping that to the true human-readable reason — and distinguishing “customer actively cancelled” from “temporary limit hit” from “ceiling set too low at signup” — is a classification/reasoning problem across messy multi-source data, and getting it wrong burns the recovery. Second, the WhatsApp recovery conversation: it’s a live multilingual (PT, with regional register) back-and-forth that has to read the customer’s replies, pick the right bank-specific instructions, handle objections, and know when the mandate is genuinely dead vs. recoverable. Strip the AI out and you’re left with a static dashboard and canned templates — which is exactly the useless thing the PSPs already ship. The AI is the product.

7. Localization angle

This is a LatAm/Brazil-only play by construction — the entire product exists because of a Brazil-specific payment rail (Pix Automático), a Brazil-specific regulatory structure (BCB mandate rules, CDC cancellation rights), and a Brazil-specific distribution channel (WhatsApp is the channel; Hotmart/Kiwify are the checkouts). Language is Portuguese-first with bank-specific instruction sets for the top Brazilian banks. Pricing is set in Reais at levels a solo creator accepts (R$97–R$397/mo), which no globally-priced churn tool bothers with. A generic global dunning product literally cannot serve this — it has no concept of a mandate ceiling. That’s the moat’s foundation.

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

  • Pricing: three tiers — Starter R$97/mo (up to ~200 active mandates), Growth R$247/mo (up to ~1,000), Pro R$397/mo (unlimited + priority WhatsApp flows). Optional success-fee variant: 15% of recovered MRR for sellers who prefer no fixed cost.
  • ACV: blended ~R$220/mo ≈ US$480/yr (at ~R$5.5/USD).
  • Rough math to $1M ARR: ~1,750 sellers × US$480/yr ≈ US$840k; ~2,100 sellers clears US$1M. Against a pool in the hundreds of thousands of recurring sellers, that’s <1% penetration.
  • Rough math to $5M ARR: ~10,500 sellers, or a smaller base with a higher-ACV Pro/agency tier (agencies managing many creators) + the success-fee upside on large sellers. Requires becoming the default recovery layer in 2–3 checkout ecosystems.
  • Expansion path: ACV grows via mandate volume tiers, then add adjacent modules — pre-emptive ceiling management, cohort churn analytics, a “reactivation” campaign engine for dead mandates, and eventually a marketplace/agency plan for consultants who run many creators’ billing.

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

  • Kiwify/Hotmart creator communities: these creators cluster in known Telegram/WhatsApp/Discord groups and Facebook groups (“Área de Membros”, “Produtores Kiwify”, etc.). Post a free “Pix Automático failure teardown” — show a real seller their failure breakdown — and convert. Target the ~29k Kiwify creators specifically.
  • Cold DM the pain: scrape sellers publicly running subscription products on Hotmart/Kiwify marketplaces, DM/WhatsApp a personalized Loom-style teardown (“aqui está por que 7% da sua recorrência falhou mês passado e como recuperar”). This is a named list + named channel.
  • Infoproduct-adjacent influencers: partner with 3–5 mid-size Brazilian creators who teach other creators how to sell (the “guru dos gurus” tier) — they have the exact audience and monetize via affiliate/rev-share. One good partner = hundreds of qualified sellers.
  • PSP/checkout co-marketing (later): Asaas, Vindi and Iugu all want merchants to succeed on Pix Automático; a recovery layer that reduces their churn is a natural listing/integration partner — warm inbound once there’s traction.

10. Build complexity — justification

Medium. Off-the-shelf: LLM APIs for classification + conversation, WhatsApp Business API (or a BSP), standard web stack, webhook ingestion. The custom work is the integration matrix — each checkout (Hotmart, Kiwify, Guru) and each PSP (Asaas, Vindi, Iugu) has its own webhook/event shape and its own way of surfacing mandate state — plus building the correct bank-specific fix instructions and keeping them current. That’s honest integration and domain work, not research. A technical founder + one helper ships a credible v1 (one checkout + one PSP + WhatsApp) in ~10–14 weeks, then expands the matrix. No novel ML, no hardware, no regulatory build.

11. Gating checklist

GatePass?Note
Legal in target market✅Sits on top of BCB-licensed PSPs; respects CDC cancellation rights (won’t re-charge cancelled mandates). No money handling.
Ethical — no harm / dark patterns✅Recovery must be honest — helps a willing customer fix a technical failure, never re-charges someone who genuinely cancelled. Guardrail this hard.
Market exists (evidence above)✅182% QoQ rail growth, card-only recovery incumbents, active creator base.
1–5 person team can build this✅Technical founder + 1; Medium complexity.
Launchable with <$50K / ₹40L✅API + WhatsApp costs are usage-based; no upfront capex.

All five pass.

12. Feasibility score

AxisWeightScoreNotes
Problem intensity2016/20Failed recurring charges = direct MRR loss, felt every cycle. Not quite hair-on-fire (they survive today with manual WhatsApp), but painful and recurring.
Demand evidence1512/15Strong indirect signals — card-recovery incumbents, explosive rail adoption, active churn-recovery content market. Docked: no verbatim seller quotes surfaced yet (must validate).
Build feasibility1511/15Off-the-shelf AI + WhatsApp, but the multi-checkout/multi-PSP integration matrix and bank-specific instruction upkeep are real work.
Distribution clarity1512/15Named communities + named seller lists + influencer tier. Conversion math still a guess; free-teardown hook is strong.
Revenue mechanics1512/15Reais pricing benchmarked to creator wallets; recovered-MRR framing makes ROI obvious. ~2,100 sellers to $1M is credible.
Time to first revenue108/10Free teardown → paid in weeks; self-serve. Gated only by shipping one checkout + one PSP integration.
Defensibility105/10Execution + accumulating mandate-failure data + bank-instruction library + checkout integrations = a real 6–12 month head start, but a PSP or a well-funded creator-tools player could copy the concept. Data/workflow lock-in is the durable part.
Total10076/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy · domain-expertise-required — needs someone who can build the integration matrix and genuinely understand Pix Automático mandate mechanics and the Brazilian creator ecosystem. Portuguese fluency and on-the-ground network are near-mandatory.

Key assumptions to validate (3–5)

  1. Assumption: A meaningful share (>4%) of Pix Automático charges fail for mandate-specific reasons that a script can actually recover. How to test: Get 5–10 sellers to share anonymized failure logs; classify reasons; measure the recoverable slice.
  2. Assumption: Sellers will pay R$97–R$397/mo when the ROI is framed as recovered MRR. How to test: Run 20 teardown → offer calls; measure close rate and price sensitivity.
  3. Assumption: WhatsApp recovery flows meaningfully out-recover the seller’s current manual/card-style attempts. How to test: A/B a small cohort — PixFiel flow vs. their status quo — over 2 billing cycles.
  4. Assumption: The checkout/PSP webhooks expose enough mandate state to diagnose failures without being the PSP. How to test: Build one Hotmart + one Asaas integration and confirm the event payloads carry ceiling/limit/cancellation signal.

Risk flags

  1. Platform dependency: Relies on Hotmart/Kiwify/PSP webhook access and WhatsApp Business API. If a checkout builds native mandate-recovery or restricts webhooks, the wedge narrows. Mitigate by spanning multiple ecosystems fast.
  2. Incumbent absorption: Asaas/Vindi/Iugu could ship “smart Pix Automático recovery” as a feature. Mitigate by owning the no-code seller segment they underserve and going multi-PSP (agnostic > any single rail).
  3. Market timing: Depends on Pix Automático failure volume being big enough now. If adoption plateaus below the pain threshold at small sellers, TAM shrinks. Adoption data says the opposite, but watch it.
  4. Compliance/ethics: CDC is strict on cancellations. Re-charging or nagging a genuinely-cancelled customer is both illegal and reputationally fatal — the classification “cancelled vs. recoverable” must be conservative.

14. Structured verdict

Score:                  76/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Technical founder, Portuguese-native, plugged into the Brazilian creator/PSP ecosystem
Time to revenue:        8–12 weeks (one checkout + one PSP + WhatsApp flow)
Capital to launch:      R$40k–R$80k / US$8–15K (mostly founder time + API/WhatsApp usage)
Top 3 assumptions to validate first:
  1. >4% of Pix Automático charges fail for recoverable mandate-specific reasons — verify with real seller logs
  2. Sellers pay R$97–R$397/mo on a recovered-MRR ROI pitch — verify with 20 teardown→offer calls
  3. Checkout/PSP webhooks expose enough mandate state to diagnose without being the PSP — verify by building one integration
Kill criteria:
  - Abandon if <4% of charges in real seller logs are recoverable in mandate terms (problem too thin)
  - Abandon if <10% of 30 teardown demos convert to paid within 60 days
  - Abandon if a major checkout (Hotmart/Kiwify) ships native mandate-recovery before v1 and closes webhook access

15. Next step — 1-week validation sprint

  • Day 1–2: Recruit 6–8 Brazilian subscription sellers already on Pix Automático (from Kiwify/Hotmart creator groups). Get anonymized failed-charge logs for the last 2 cycles.
  • Day 3–4: Manually classify every failure into mandate-specific buckets (ceiling-too-low, limit-hit, cancelled-in-bank, insufficient-funds). Compute the recoverable percentage. Hand-run 10–15 WhatsApp recovery attempts using bank-specific scripts and measure how many revive.
  • Day 5: Decide go / no-go. Falsifiable bar: ≥4% of charges are recoverable-in-mandate-terms AND ≥3 of 8 sellers say “yes, I’d pay R$150+/mo for this” after seeing their own recovered-R$ number. Below either line → no-go or rework the wedge.

Interested in a detailed proposal?

Get a deep-dive with market research, competitive analysis, and implementation roadmap.

Contact us

info@startupbasket.ai