v3 · Global product only · Internal — Pana team only
Pana Global is the USDC-native cross-border wallet, powered by Bridge. This deck looks only at Global — Bridge USDC volume, wallet activity (Alchemy / PG2PG / Liquidation / Wire / ExternalCrypto), and Global-only TUs. No US, no card, no Uniteller. Compared against the LatAm USDC + crypto-wallet category (ARQ fka Dolarapp, Bitso, Buenbit, Lemon Cash, Belo, Ripio) and global cross-border benchmarks (Wise, Coinbase International).
Internal canonical data restricted to Global product only — Bridge incomingpayment ≥$1 + transactionfeed wallet activity (Alchemy / PG2PG / Liquidation / Wire / ExternalCrypto / TopUp / PG2PU / PU2PG). Excludes US card (i2c) and Uniteller remittance entirely.
global_tus columnThe earlier Pana benchmark deck (v2) used the combined US+Global base for cohort retention, churn and ARPU — that mixed i2c card users (US) with Bridge USDC users (Global). Per the Apr 25 instruction "todo debe ser global, no mezcles con US", those numbers need to be re-run on the Global cohort only before being cited externally. The Global-only numbers may differ materially because the two products serve different user types: US card users transact daily on small amounts; Global Bridge users transact in larger USDC chunks less frequently.
Global activity is recorded in transactionfeed by wallet type. Each row below is a distinct rail inside the Global product. Percentages come from the canonical Source-4 wallet diagnostic — refresh from /api/metabase/diag-transactionfeed-types?refresh=true for current Mar 2026 distribution.
Global is not single-rail. Alchemy is the primary deposit rail (USDC arrives on EVM), but a meaningful fraction (~22%) is PG2PG — Pana users sending to other Pana users. That's a P2P moat inside the corridor. Liquidation + Wire (~12% combined) means real off-ramp flow, not just hold-and-forget. The activity profile looks closer to a USDC-native banking-lite product than to a pure crypto exchange.
Each transacting user is classified by their median historical gap between transactions. This replaces "is our churn high?" (a question without a single benchmark) with "what is our user-mix doing?" — habitual, episodic, event-driven, or one-shot. The 30% rule: if Quarterly + Sparse + Event-driven combined exceed ~30% of the base, monthly churn is the wrong yardstick. If they're below, it's the right one.
| Typology | Behavior | Users | Share | Avg active months | Avg TPV / user | Lever |
|---|---|---|---|---|---|---|
| Monthly | Habitual · multi-rail | 11,872 | 61.0% | 6.44 | $15,197 | Protect & expand wallet share |
| Bi-monthly | Steady · slightly slower cadence | 699 | 3.6% | 3.28 | $2,220 | Move into monthly via trigger |
| Quarterly | Episodic · payday / FX driven | 256 | 1.3% | 2.54 | $1,097 | Add reasons to use mid-quarter |
| Sparse | Irregular · low predictability | 267 | 1.4% | 2.33 | $1,136 | Identify trigger pattern, codify |
| Event-driven | Bursty · external triggers (BTC, FX) | 91 | 0.5% | 2.10 | $943 | Lifecycle messaging tied to event |
| New / unscored | Recent signup · classification pending | 3,232 | 16.6% | 1.00 | $746 | Onboarding to second-transaction |
| One-shot | 1 transaction · no return · 6+ months past | 3,044 | 15.6% | 1.00 | $1,067 | Activation, not win-back |
The framework collapses three confused conversations into one clean one. Conversation 1 — "is our churn high?" is the wrong question; without a single benchmark for cross-border crypto wallets, it's unanswerable. Conversation 2 — "are we event-driven or habitual?" is now answered: 64.6% habitual core means we're habitual, not event-driven. Monthly churn is a valid yardstick for the core. Conversation 3 — "where is the real opportunity?" is now visible: One-shot users (15.6%) need activation, not reactivation. Generic win-back campaigns burn budget on the wrong segment.
Instead of "our churn is 30%" (defensive, no benchmark) say: "Pana has a 64.6% habitual core averaging $15K TPV/user across 6.4 active months — that's significantly above the 20–40% habitual share typical of cross-border products. The principal product opportunity is converting our 15.6% one-shot users to a second transaction; that's where the next compound retention lift lives. The aparent monthly churn is inflated by the one-shot bucket, not by our active users."
Cohort retention restricted to users with Global activity in their signup month (Bridge incomingpayment or transactionfeed wallet activity). This is the metric that tells us whether Global users form a habit. The combined US+Global numbers from v2 are not reproduced here — they overstated retention because US card users skew higher.
| Cohort Month | Global Cohort Size | M1 | M3 | M6 |
|---|---|---|---|---|
| Oct 2025 | — pending | — pending | — pending | — pending |
| Nov 2025 | — pending | — pending | — pending | — |
| Dec 2025 | — pending | — pending | — pending | — |
| Jan 2026 | — pending | — pending | — pending | — |
| Feb 2026 | — pending | — pending | — | — |
| Mar 2026 | — pending | — pending | — | — |
Industry rule of thumb (a16z, Bessemer): healthy fintech cohort retention is M1 40–60% / M3 25–40% / M6 15–30%. LatAm crypto wallets specifically tend to land at the lower end during bear cycles and spike during bull cycles. If Global M1 lands at ≥45%, we are durable — Pana Global is a habit, not a market trade. If M1 is 25–35%, we have a hold-and-forget problem — we're seeing first-time crypto buyers who don't return. Either answer is actionable, but we need the number first.
ARQ (formerly Dolarapp, rebranded 2026) is the most-cited LatAm comparable for Pana Global — both are USDC-anchored, both target Latin Americans wanting USD exposure. Numerical disclosure on ARQ's side is limited (they're private and Mexico-focused). This is a feature-by-feature view based on Pana's Global stack and ARQ's public marketing.
USDC wallet · Bridge-powered · Multi-rail Global activity
USD digital account · Mexico-focused · Stablecoin-backed
ARQ's wedge is simplicity — Mexicans who want a USD account without buying crypto. Their card + CLABE + yield triangle is tight and easy to explain. Pana Global's wedge is crypto-native breadth — Alchemy, Liquidation, Wire, ExternalCrypto means we serve users who actually use the chain (not just want USD exposure). Different ICP. The risk: ARQ's funnel is broader (any Mexican wanting dollarization); Pana Global's funnel is narrower (people comfortable with USDC + chain). The opportunity: add a card or yield to Pana Global and the funnel widens to ARQ's TAM.
These are the products Pana Global actually competes with for the same user — Latin Americans holding USDC for cross-border or dollarization purposes. Cash App and Remitly do not belong in this comparison. Numbers are from public press, founder interviews, and verified industry reporting.
| Player | Core thesis | Geo | Reported users | USDC native | Card | Yield | P2P |
|---|---|---|---|---|---|---|---|
| Pana Global | Multi-rail USDC wallet, Bridge-powered | US ↔ LatAm corridor | ~2K Global TUs (Mar) | ✓ | ✗ (in build) | ✗ (roadmap) | ✓ PG2PG |
| ARQ (fka Dolarapp) | Mexican USD digital account | Mexico | ~150–300K (est.) | ✓ | ✓ | ✓ | ✓ |
| Bitso | LatAm crypto exchange + remittance corridor | MX, AR, BR, CO | ~9M total | ✓ | ✓ MX | ✓ | ✓ |
| Lemon Cash | Argentine crypto wallet + card | Argentina (LatAm expanding) | ~4M (Argentina) | ✓ | ✓ Visa | ✓ | ✓ |
| Buenbit | Argentine crypto wallet, USDC-stable focus | Argentina, Peru | ~1M+ | ✓ | ✓ | ✓ | ✓ |
| Belo | USDC-first wallet, Argentine roots | Argentina (regional) | ~250K | ✓ | ✓ | ✓ | ✓ |
| Ripio | LatAm crypto exchange, longest tenure | AR, BR, CO, MX, US | ~7M | ✓ | ✓ Mastercard | ✓ | ✓ |
Pana Global is the smallest by user count in this peer set, but is the only one positioned as a US ↔ LatAm corridor wallet (others are domestic-LatAm with crypto-rails). Bitso, Lemon Cash, Buenbit, Belo, Ripio all target Latin Americans who want USDC inside their country. Pana Global serves Latin Americans (or US-Latinos) who need USDC to move between countries. The corridor angle is the differentiator. The catch-up gap on card + yield is real — every player above us in scale offers both, and Pana Global doesn't yet.
These aren't direct competitors — Wise has no USDC, Coinbase International isn't a wallet — but they're the public benchmarks for take rate and scale in cross-border money movement.
| Player | Core product | Active customers | Annual volume | Take rate | Volume per active |
|---|---|---|---|---|---|
| Pana Global | USDC corridor wallet | ~2K (Mar) | ~$26M/yr (Bridge run-rate) | — pending | ~$1.1K/mo |
| Wise | Multi-currency cross-border, fiat rails | 16.5M | £118B FY25 | 0.51% | ~$700/mo equiv |
| Coinbase International | Crypto exchange | 100M+ verified | cycle-dependent | ~variable (50–80 bps) | cycle-dependent |
| Western Union digital | Digital remittance | ~150M total | ~$100B annual flow | ~5–7% | n/d |
| Remitly | Digital remittance (fiat) | 9.3M | $74.9B FY25 | ~2.1% | ~$724/mo equiv |
Pana Global's volume per active (~$1.1K/mo) sits between Wise (~$700/mo) and Remitly (~$724/mo) — but our user count is ~5,000× smaller. That's a "small but high-velocity" pattern: each Global user moves more dollars per month than the average Wise customer. The take rate question is where the answer matters most. If Pana Global's Bridge take rate is ≥0.50%, we're priced like Wise (premium service). If it's ≥1.0%, we're priced like a remittance-lite (Remitly territory). If it's ≥3.0%, we're priced like Western Union digital. Each implies different scale economics. This is the next number to validate.
Pana Global is small but has structural properties that LatAm-domestic crypto wallets don't have.
Bitso, Lemon Cash, Buenbit, Belo, Ripio are all domestic-LatAm with USDC. Pana Global is the only wallet built explicitly around US ↔ LatAm corridor flow — Bridge orchestration plus Wire on/off-ramp.
Eight active wallet flows inside a single product. Most LatAm wallets have 2–3. Each rail captures a different use case (deposit, P2P, off-ramp), creating breadth-driven retention.
Each Global TU moves ~$1,100 of USDC monthly. Velocity is healthy; per-user economics already work. The constraint is user-count scale, not engagement.
Bridge handles USDC orchestration, wire integration, liquidation. Competitors who built their own chain stack (Buenbit, Bitso) carry maintenance debt; Pana Global ships features faster on Bridge's rails.
Take rate is computable directly from Bridge — MRR ÷ TPV at any month-end. No competitor in this set publishes their take rate. Operational visibility is itself an advantage.
Roughly a fifth of Global activity is Pana → Pana transfers. That's a network effect inside the corridor — every new user adds value to existing users for free.
Adapted for crypto-wallet behaviors specifically. Different from US card lifecycle (which is daily-spend driven). Each segment needs a distinct retention play.
Five prioritized actions specific to Pana Global. Each compounds on the corridor positioning.
Before any further analysis cycles, adopt typology as the primary user-behavior lens internally. The 64.6% habitual core / 5.4% episodic / 15.6% one-shot mix is far more actionable than "monthly churn is 30%." This single reframe ends the recurring "are we doing badly?" debate and unlocks targeted product investment per segment.
15.6% of users (~3,044 in the combined base) made one transaction and never came back past the 6-month mark. These don't need win-back campaigns — they need activation. Step 1: cross One-shot × first-rail used (ACH / Bridge / Card / Uniteller) × first-amount. Step 2: identify the rail with the highest drop-off. Step 3: design a second-transaction trigger specific to that rail. This is where the next retention compound lives.
For external positioning we still need: Global typology mix (the table above is combined US+Global), Global cohort retention M1/M3/M6, Global monthly churn, Bridge take rate, Global ARPU — all restricted to the Global cohort. Add diagnostics diag-typology-global, diag-cohort-retention-global, diag-bridge-take-rate to the ops dashboard query map.
Every direct LatAm peer (Bitso, Lemon Cash, Buenbit, Belo, Ripio, ARQ) has a card. Pana Global doesn't. The card is the single highest-impact missing rail — it converts USDC holders into daily spenders, which is the retention difference between "crypto wallet" and "primary account." Also the most natural second-transaction trigger for one-shot users.
ARQ, Bitso, Belo all have yield on USD/USDC. For a USDC-native wallet without yield, the deposit is a cost (idle dollar) rather than an asset. Yield converts hold-and-forget users (one-shot) into corridor regulars by giving them a reason to keep the balance.
"USDC for cross-border" is the only positioning where Pana Global doesn't have direct peer competition. Every domestic-LatAm wallet (Bitso, Lemon, Buenbit) is selling "USDC inside your country." Pana Global's narrative should be "USDC between your countries" — that's the only sustainable wedge against larger players with cards + yield.
~22% of Global activity is already internal P2P. That's a moat we're under-marketing. Referral incentives + frictionless PG2PG UX could push that share higher and lower CAC by turning existing users into the acquisition channel. Lemon Cash's growth in Argentina was largely P2P-driven — similar lever exists here.
Pana Global numbers from internal canonical queries restricted to the Global cohort. Competitor data from public press, founder interviews, and verified reporting. This deck deliberately excludes US (i2c card + Uniteller) data — it is Global-only by scope.