---
title: "How to evaluate a global checkout provider: 12 questions"
seoTitle: "How to evaluate a global checkout provider: 12 questions"
description: "A 2026 scorecard for choosing a global checkout provider: 12 questions on rails, pricing, sandbox, webhooks, compliance, settlement, and custody."
date: "2026-10-29"
updated: "2026-10-29"
category: "payments"
author: "BlindPay Team"
faq:
  - q: "What should I look for in a global checkout provider?"
    a: "Look for named local rails per country, a fee you can see in a live quote before the shopper pays, a sandbox that can force failures, idempotent money calls, signed webhooks for every status, compliance run inside the API, a clear settlement time per rail, and a clear answer on who holds the funds. Score each on a 0 to 2 scale and weight the ones your first market depends on."
  - q: "How do I compare payment method coverage across providers?"
    a: "Ask for a rail map by country, not a country count. A provider that lists 100 countries may reach most of them through international cards or SWIFT. Ask which rail a Brazilian, Mexican, or Colombian shopper actually uses, what limits apply per shopper, and whether the merchant needs a local entity for that rail. Then test your top two corridors end to end."
  - q: "Why does seeing the fee before payment matter at checkout?"
    a: "Because extra costs are the top reason shoppers leave. Baymard counts extra costs too high at 40% of US shoppers who abandon. A provider that returns a live quote with the local amount, fees, and expiry lets you show the shopper the exact total and the merchant the exact amount received, instead of reconciling surprises after settlement."
  - q: "How do I test a checkout provider's sandbox?"
    a: "Run the full loop: onboard a test merchant, create a session, pay on each rail, and force a failed and a refunded payment. On BlindPay development instances, payins complete by themselves in about 30 seconds, request_amount 66600 forces failed, and 77700 forces refunded. Check that every status reaches your webhook handler and that a retried request with the same idempotency key does not create a second payment."
  - q: "What are red flags when choosing a checkout provider?"
    a: "Coverage counts with no rail names, fees that appear only on the settlement report, a sandbox that only returns success, no idempotency on payment creation, unsigned webhooks, onboarding times nobody will state, and vague answers on who holds the funds before payout. Any one of these turns into a support ticket or a reconciliation gap in your first month live."
---

To evaluate a global checkout provider, score it on 12 questions across six areas: rail coverage per country, fee visibility, developer experience, compliance, settlement and failure handling, and custody. Ask each provider the same questions, score every answer 0 to 2, and test your top two corridors end to end before you sign anything.

This scorecard reflects public information as of October 2026. For the build itself, read the [global checkout checklist](/resources/more/global-checkout-checklist). For the basics, start with [what is a global payments checkout](/resources/more/what-is-a-global-payments-checkout).

**Key takeaways**

- Ask for a rail map by country. A country count hides whether shoppers pay on Pix, SPEI, or PSE or on an international card.
- Extra costs are the top abandonment reason in Baymard's research at 40%, so the fee has to be visible in a quote before the shopper pays.
- A good sandbox forces failures. On BlindPay development instances, 66600 forces a failed payin and 77700 a refunded one.
- KYB time sets your merchant activation time. BlindPay KYB Standard takes 3 hours to 1 business day.
- Know who holds the funds between payment and payout, and which rails pay them out.

## How should you run the evaluation?

Score answers, not demos. Send every provider the same 12 questions, then verify the best two in a sandbox.

1. **Write down your first two markets** and the rails shoppers there use.
2. **Send the 12 questions below** to each provider, in writing.
3. **Score each answer** 0 (no answer or vague), 1 (answered, with gaps), or 2 (specific, documented, testable).
4. **Weight the questions** your first market depends on. A Brazil launch weights Pix questions double.
5. **Test the top two** in their sandboxes, running the same checkout on both.
6. **Decide on the total**, with a hard fail for any 0 on compliance or custody.

## What are the 12 questions?

Twelve questions, grouped into six areas. Copy the table into a spreadsheet and fill one column per provider.

| # | Question | Why it matters | A good answer | Score 0 to 2 |
| --- | --- | --- | --- | --- |
| 1 | Which rail does a shopper use in each of my markets? | Shoppers convert on the method they know | Named rails per country: Pix, SPEI, PSE, Transfers 3.0, ACH, RTP | |
| 2 | Does the merchant need a local entity for each rail? | Entity setup can add months | Rails available to a foreign merchant, stated per country | |
| 3 | What limits apply per shopper and per transaction? | Limits cap your average order value | Documented minimums and maximums per currency, with the error you get | |
| 4 | Is the fee in a live quote before the shopper pays? | Surprise costs drive abandonment | A quote with local amount, fees, rate, and expiry | |
| 5 | Can I price in one currency and receive an exact amount? | Merchants reconcile against list price | A receiver-denominated quote where the merchant gets the full price | |
| 6 | How does a platform earn a fee on each payment? | Platforms need a take rate | A per-transaction fee field on the quote, with a payout schedule | |
| 7 | Can the sandbox force failed and refunded payments? | Your error paths need testing too | Documented test amounts for each outcome | |
| 8 | Are payment-creating calls idempotent? | Retries must not double charge | An idempotency key that replays the original response | |
| 9 | Which webhooks fire, and are they signed? | Status drives fulfillment | Created, updated, and finished events, signed and verifiable | |
| 10 | Who runs KYB, KYC, and screening, and how long does it take? | Activation time and compliance burden | Checks inside the API, with stated review times | |
| 11 | When do funds settle, and how do failures and holds show up? | Cash flow and support load | Settlement time per rail, plus explicit failed, refunded, and on hold statuses | |
| 12 | Who holds the funds before payout, and how do they leave? | Counterparty risk and payout reach | A clear custody model and named payout rails | |

The coverage questions come first because they decide conversion. Stripe's [test of 50+ payment methods](https://stripe.com/blog/testing-the-conversion-impact-of-50-plus-global-payment-methods) found that surfacing at least one relevant method beyond cards lifted conversion 7.4% and revenue 12% on average, with results varying by method, country, and industry. [Do local payment methods raise checkout conversion?](/resources/more/local-payment-methods-checkout-conversion) breaks those numbers down.

## What does a strong answer look like on coverage and limits?

A strong answer names the rail, the currency, and the limit, and you can check all three in the docs.

Ask for specifics, because they differ by provider. Stripe's [Pix docs](https://docs.stripe.com/payments/pix), for example, state settlement of up to 2 business days and limits of US$3,000 per day and US$10,000 per month for each Brazilian customer paying a business outside Brazil. That is the level of detail to expect from every provider on every rail. If an answer stops at "we support Brazil", score it 0.

## What does a strong answer look like on pricing?

The fee shows up in a quote before the shopper commits, and the merchant knows exactly what lands.

[Baymard's abandonment research](https://baymard.com/lists/cart-abandonment-rate) lists "extra costs too high" at 40% and "couldn't see or calculate total order cost up front" at 12% among US shoppers who left a cart. A provider that returns a live quote lets you show one total in the shopper's currency. Stripe's [localization guide](https://stripe.com/resources/more/payment-localization-best-practices-a-guide-for-scaling-payments-worldwide) also recommends showing and charging in the buyer's currency through the final charge.

## What does a strong answer look like on compliance and custody?

Checks run inside the payment API with stated times, and the provider can say in one sentence who holds the money.

Compliance answers decide activation time. If merchants wait a week for KYB, your sales team waits a week too. Custody answers decide risk: whose name the balance sits in between the shopper's payment and the merchant's withdrawal, and what happens to a payment that cannot complete. Score a vague answer on either as a hard fail. For a deeper vendor diligence process, see [the crypto payment processor scorecard](/resources/more/crypto-payment-processor-for-businesses).

## How does BlindPay answer the 12 questions?

BlindPay is the money layer you build a checkout on. Here is how it answers, from its public docs.

1. **Rails:** Pix (Brazil), SPEI (Mexico), PSE (Colombia), Transfers 3.0 (Argentina), ACH, wire, RTP, and ACH pull (US).
2. **Entity:** Each merchant is a BlindPay business customer that passes KYB.
3. **Limits:** The quote enforces min and max per currency and names the range in its error.
4. **Fees:** A live payin quote with the local amount, fees, and an `expires_at` 5 minutes out. Pricing is a monthly plan plus per-transaction fees disclosed in the quote.
5. **Exact amount:** `currency_type: "receiver"` with `cover_fees: true` delivers the full USD price.
6. **Take rate:** `partner_fee_id` on the quote, paid out monthly.
7. **Sandbox:** 66600 forces failed, 77700 forces refunded.
8. **Idempotency:** `Idempotency-Key` replays the original response.
9. **Webhooks:** `payin.new`, `payin.update`, `payin.complete`, signed.
10. **Compliance:** KYB Standard 3 hours to 1 business day; screening in the flow.
11. **Settlement:** Pix, SPEI, Transfers 3.0, and PSE typically minutes; `on_hold`, `failed`, and `refunded` statuses.
12. **Custody and payouts:** Managed wallets (custodial, beta) or a wallet you control; payouts over Pix, SPEI, SEPA, Faster Payments, and SWIFT (POBO/COBO).

## Methodology and sources

BlindPay facts come from the [payins](/docs/payins), [payin quotes](/docs/payin-quotes), [partner fees](/docs/learn/partner-fees), [idempotency](/docs/learn/idempotency), [webhooks](/docs/learn/webhooks), and [cut-off times](/docs/kb/cut-off-times) docs, read in October 2026. External sources: [Stripe's 50+ payment methods test](https://stripe.com/blog/testing-the-conversion-impact-of-50-plus-global-payment-methods), [Stripe's Pix docs](https://docs.stripe.com/payments/pix), [Stripe's localization guide](https://stripe.com/resources/more/payment-localization-best-practices-a-guide-for-scaling-payments-worldwide), and [Baymard's cart abandonment research](https://baymard.com/lists/cart-abandonment-rate). Provider capabilities change, so ask each one to answer in writing with a date.

## What to do next

Copy the 12-question table into a spreadsheet, add one column per provider, and send it out this week. Score the answers, then run the same checkout in the top two sandboxes, including one forced failure, before you sign.