International card success rate measures how often a defined set of cross-border card-payment attempts reaches a specified successful outcome. To compare Razorpay with another payment option fairly, use the same outcome, attempt definition, time period and customer segments. A higher percentage alone does not establish better collections or lower total cost.
One provider quotes 92% and another quotes 87%. Before deciding which result is better, ask what each number counts. An issuer-approval rate, a captured-payment rate and the proportion of customers who eventually pay are different measures. This guide shows how to read them, identify avoidable failures and turn the results into a useful business comparison.
Key Takeaways
- Define exactly what a successful payment means. Authorised, captured and settled payments are different outcomes; state both the numerator and denominator.
- Report first-attempt approval and later recovery separately. Distinguish individual attempts from unique-order conversion and recovery after retries.
- Compare the same customer and transaction segments. Match issuing countries, card types, recurring-payment mix and reporting window.
- Track payment value and risk outcomes alongside payment counts. A count-based approval rate does not automatically describe captured revenue. Review fraud, disputes and refunds as well.
- Use your account’s evidence to evaluate Razorpay. Provider-published success claims are not a guaranteed result for every business.
What Does International Card Success Rate Actually Measure?
A useful success rate specifies the payment event it measures and the population included. The general calculation is successful outcomes divided by eligible attempts, multiplied by 100. The reporting definition determines what belongs in each part of that calculation.
Success rate = successful outcomes ÷ eligible attempts × 100
If the question is issuer approval, count submitted authorisation requests and approvals. If the question is completed sales, measure unique orders and captured payments instead. Do not treat every API call as a new attempt: a status lookup, duplicate request or other non-authorisation action is not automatically another request for issuer approval.
Attempt-based, order-based and customer-based reporting
| Measure | Population | Useful question |
|---|---|---|
| Attempt-based approval | Submitted authorisation requests under a fixed reporting definition. | How often did the issuer approve the requests it received? |
| Unique-order conversion | Eligible orders, deduplicated across attempts. | How many orders eventually produced a captured payment? |
| Customer-level completion | A defined set of customers over a stated period. | How many customers eventually paid through an eligible method? |
Suppose one customer makes four genuine card authorisation attempts for the same order and the fourth succeeds and is captured. Attempt-based approval is 1 ÷ 4, or 25%. Order-based completion is 1 ÷ 1, or 100%. Both can be accurate when clearly labelled, but they answer different questions.
Did You Know? A retry can improve the chance of collecting an order while lowering an attempt-based percentage if it adds several unsuccessful requests. Report recovery separately so this does not hide either the customer outcome or the processing effort.
Authentication, Authorisation, Capture and Settlement: What Is the Difference?
These measures describe different stages of a payment. A customer can complete authentication but still receive an issuer decline. An approved payment can remain uncaptured, and a captured payment can await its scheduled settlement.
| Stage | What to measure | What it does not establish |
|---|---|---|
| Authentication | Completion of the applicable authentication flow, such as a 3DS challenge. | That the issuer approved the purchase. |
| Authorisation | Issuer approvals among submitted authorisation requests. | That the payment was captured or settled. |
| Capture | Captured payments among the eligible authorised payments. | That the bank settlement has already completed. |
| Settlement | Processing or bank credit under a clearly stated settlement definition. | The original checkout conversion rate or final profit. |
| Order completion | Eligible unique orders with a captured payment. | That every attempt for those orders succeeded. |
When reviewing a payment gateway, ask for the reporting definition as well as the percentage. Razorpay’s settlement documentation separately requires captured payments and an activated account for settlement.
Why Can International Card Performance Differ from Domestic Performance?
Issuing country, card type, customer behaviour, authentication, merchant risk profile and acquiring arrangements can differ across domestic and international traffic. Those differences can change the result even when the website and payment amount appear similar.
There is no universal percentage-point gap that every Indian exporter should expect. A provider’s portfolio benchmark is useful only when its dates, market, transaction type and calculation match the question you are asking.
Use the cardholder’s relevant segment
An Indian business accepting foreign-issued cards should investigate those foreign issuers and cardholders. An authentication ranking for cards issued in India does not, by itself, explain the experience of its overseas customers. Segment by issuing country rather than assuming the merchant’s location represents the entire population.
Separate recurring and one-time payments
A first purchase and a recurring subscription renewal can have different credentials, consent requirements and failure reasons. Report them separately before judging a change in the overall rate. Also check card brand, debit versus credit, device and the authentication flow where the data is available.
Pro Tip: Start with your three largest international issuing-country segments. For each, review first-attempt approval, order completion and captured value. This is more actionable than changing the entire checkout after a movement in one blended number.
What Can a Blended Percentage Hide?
A blended rate is a weighted result across its included transactions. It can change because performance improved, because the mix changed, or both. Separate these explanations before attributing a gain to a provider or a product feature.
For example, assume a period contains 800 domestic attempts at 93% approval and 300 cross-border attempts at 71%. There are 744 domestic approvals and 213 cross-border approvals: 957 out of 1,100 attempts, or 87%. This is an illustration of weighting, not an observed provider benchmark.
If the next period contains fewer cross-border attempts, the blended percentage can rise even when neither segment improves. Similarly, removing risk-blocked traffic from the denominator may make issuer approval look higher without improving completion across all eligible orders.
Track checkout conversion separately
A gateway approval measure generally starts later than a website-visitor or checkout-session measure. Keep the stages separate, but examine their relationship. Better authentication or fewer avoidable declines may improve completed sales; a confusing checkout can prevent customers reaching the authorisation stage at all.
A practical report follows the journey from eligible checkout sessions to submitted attempts, approvals, captured orders and subsequent refunds or disputes. Define each stage rather than relying on one label for the entire journey.
How Should You Compare Razorpay and Other Providers Fairly?
Use a matched cohort and the same measurement rules. If the available results cover different countries, ticket sizes, payment types or dates, explain the difference rather than presenting the percentages as a controlled comparison.
- Fix the outcome. State whether success means issuer approval, capture or eventual order completion.
- Fix the denominator. Specify eligible requests and whether pre-authorisation blocks, authentication failures and retries are included.
- Use the same period. Include relevant outage or sale-event days; show daily results where a monthly average hides a change.
- Match issuing-country and card mix. Separate major card brands and relevant card types.
- Separate payment types. Do not blend recurring renewals with one-time purchases without showing both.
- Match ticket sizes. Review comparable amount bands and report value-based results.
- Control customer and order duplication. Use stable identifiers to distinguish attempts from completed orders.
- Show recovery. Report first-attempt approval and later successful recovery separately.
- Show risk outcomes. Include applicable fraud, dispute and refund measures for the same cohort.
- Explain uncertainty. Retain sample sizes, exclusions, feature changes and account-specific conditions.
If a provider cannot immediately supply a field, request a sample export or clarification. That is a reporting question to resolve; it does not prove the provider has never audited its performance or lacks the capability.
How Should You Interpret Razorpay’s International Success-Rate Claims?
Razorpay’s international-payments page describes intelligent routing, smart retries and an in-house card switch, alongside advertised success-rate outcomes. Treat these as provider-reported product claims, rather than a guaranteed approval rate for every merchant.
Before using an advertised figure in a business case, request its measurement period, transaction cohort, success definition and baseline. If it describes an uplift, ask whether that means a relative percentage increase or a percentage-point change. Confirm which features are enabled for your account and collection flow.
For your own comparison, use the payments actually processed for your business. Where a reporting view does not expose a particular segment, confirm which Dashboard report, transaction export or supported API fields can provide it. Do not assume every analytical view is included for every product.
Did You Know? Moving from 80% to 84% is a four-percentage-point increase and a 5% relative increase. A performance claim should say which calculation it uses.
Which Changes Can Help Reduce Avoidable Failures?
Investigate the recorded failure stage before choosing a change. Authentication drop-offs, issuer declines, integration errors and missed captures need different fixes. Test the effect on a defined segment rather than assuming every failed payment has the same cause.
Routing and acquiring arrangements
Routing can help address differences in acquiring performance where the merchant’s product and contractual arrangements support it. An Indian merchant should not assume that choosing an acquirer in another country automatically turns a cross-border sale into a domestic transaction. Merchant location, scheme rules and the actual acquiring arrangement matter.
If considering Razorpay’s Optimizer routing product, assess its supported integrations, commercial terms and activation separately. Mentioning a routing feature does not establish that every international merchant already has the same configuration.
Credentials and tokenisation
Supported tokenisation and credential-lifecycle updates can reduce some credential-related failures. They do not guarantee that every reissued card stays usable or override issuer decisions, customer consent or account status. Check the applicable support for the card, country and payment flow.
Retries and customer action
Use the failure code and the current scheme/acquirer rules to decide whether to retry. A temporary issue may permit another attempt; an invalid or closed account, revoked recurring permission or a prohibited retry requires different action. Some failures need the customer to update details or choose another method.
A single blanket retry limit is not a complete rulebook. Confirm the relevant response category, permitted timing and limits with the acquirer. Track duplicate-charge prevention and the eventual order outcome alongside recovery.
Is a Higher Approval Rate Always Better?
A higher approval rate is useful when it produces more legitimate completed sales at an acceptable cost and risk level. It is less informative when its denominator changed, captured value did not improve, or subsequent disputes and refunds increased.
Compare payment-level outcomes using consistent fraud and dispute definitions. Ask the acquirer which monitoring programme, geography, effective date, ratio and minimum event count apply to the account. Do not apply a generic Visa threshold to every Indian merchant or treat an ordinary chargeback percentage as the complete VAMP calculation.
Pro Tip: Put approval, captured value, refunds, disputes and processing costs in the same review. Where the populations differ, label them. This makes the operational trade-off visible without treating a single approval figure as the full result.
What Should an Indian Business Verify Before Activating International Cards?
Confirm business eligibility, supported methods, required website policies, account KYC and banking-partner approval. Razorpay’s international debit and credit card documentation describes special security and risk checks and the activation request process.
Regulatory scope also follows the collection flow. RBI’s PA-CB framework includes an exception where card networks facilitate FX settlement and the aggregator receives local currency, and banks acting in that role do not need separate PA authorisation. Verify the provider’s entity and applicable activities rather than assuming a single licence rule describes every card transaction.
Settlement currency, export documentation and export realisation obligations remain separate checks. A successful authorisation does not establish that a product supports a particular EEFC arrangement or outbound overseas payment.
How Do You Translate Success Rate into Collections?
Use captured value to assess collections. Multiplying total attempted value by a count-based success rate is valid only under suitable assumptions, such as equal ticket sizes and successful capture after approval, or when the rate itself is value-weighted.
The following hypothetical example assumes equal tickets, the same ₹10 lakh attempted cohort, successful capture after approval, no later recovery, and no refunds or disputes. It does not compare real providers.
| Illustrative rate | Captured value under these assumptions | Not captured in the window |
|---|---|---|
| 80% | ₹8,00,000 | ₹2,00,000 |
| 85% | ₹8,50,000 | ₹1,50,000 |
| 90% | ₹9,00,000 | ₹1,00,000 |
For a real business case, sum actual captured payments and deduct the relevant fees, FX costs, tax and adjustments. Separate subsequently recovered orders from genuinely uncollected orders. Net collected receipts are not profit: fulfilment costs, margins and other business expenses still matter.
A reusable comparison report should show attempted count and value, first-attempt approvals, captured orders and value, post-retry recovery, fees and relevant later risk outcomes. It should also state its reporting date so unresolved captures and disputes are not mistaken for final results.
Frequently Asked Questions
Is 90% a good international card success rate?
It depends on the definition and cohort. A rate for issuer approvals on one-time payments is not directly comparable with end-to-end completion for recurring orders. Compare matched issuing countries, card mix, dates and risk outcomes before calling a figure strong or weak.
Should retries count as separate attempts?
For an attempt-based measure, genuine submitted authorisation requests can count separately under the reporting definition. For unique-order completion, deduplicate them to the order. Publish first-attempt approval and later recovery separately.
Does a high approval rate prove Razorpay will collect more revenue for my business?
A public claim does not prove that account-specific outcome. Compare the same eligible cohort using actual captured value, declared fee and FX assumptions, and relevant risk outcomes. Confirm which features are active for your business.
Should I track success by count or value?
Track both. A business can have many successful small payments while larger orders fail. Count-based and value-based rates expose different parts of that pattern.
How often should I review the report?
Use a regular operational review suited to your payment volume, with additional checks after integration changes, sale events or unusual declines. Keep a stable reporting definition and compare both daily patterns and an adequately sized longer window.
