Razorpay vs PayU Payment Analytics: Diagnosing Checkout Failures and Closing the Ledger
The payments dashboard says success rate fell from 78% to 71% this week. Finance says bank credits for the month are ₹3.2 lakh short of what the sales report shows. Two different problems are hiding in one bad week: a checkout-performance problem and a reconciliation problem. They need different data, different people and different reports.
This guide compares how Razorpay and PayU help you separate and solve those two problems: diagnosing why payments fail, then matching sales, fees, refunds and settlements to the bank so you can close the books. It does not cover settlement speed, pricing or B2B collections; see our instant settlement comparison for timing questions.
Facts come from each provider’s dashboard and reporting documentation, checked on 12 November 2026. Report fields and limits change; confirm them in your own dashboard before building a process around them.
Quick answer: Razorpay’s Analytics Suite is built for diagnosing checkout performance: date-range comparison, payment-method filters, payment flow views and failure analysis, with gateway-level views for Optimizer merchants. Its Custom Reports let you combine fields from payments, orders, refunds and settlements into one export for reconciliation. PayU’s dashboard provides Business Summary, Transactions and Settlements views, with transaction, settlement, refund and monthly TDR reports that can be downloaded or scheduled. For teams that need to diagnose failures and close the ledger from the same provider, Razorpay’s combination of analytics and multi-entity custom reports is the stronger fit; PayU’s scheduled reports suit teams that mainly need regular files delivered to finance.
Key takeaways
- Treat success-rate drops and reconciliation gaps as separate investigations with separate data.
- Fix the date range, time zone, filters and denominator before comparing success rates; count orders and attempts separately.
- Razorpay Analytics data starts from when Analytics is enabled on the account; gateway filters apply to Optimizer merchants.
- Razorpay Custom Reports combine fields across entities so one export can carry payment ID, order ID, fee, tax, refund and settlement ID together.
- PayU documents CSV and XLSX reports with daily, weekly or monthly scheduling to up to two recipient emails per scheduled report.
Why do orders, attempts, captures and settlements need separate definitions?
Most analytics arguments are definition arguments. Agree on these first:
- Order: one purchase intent, such as a cart checkout.
- Attempt (payment): each try to pay that order. One order can have several attempts.
- Authorisation: the issuer approves the amount.
- Capture: the merchant claims the authorised amount; only captured payments become revenue.
- Refund: money returned on a captured payment, fully or partly.
- Settlement: money the gateway sends to your bank, net of fees, taxes, refunds and adjustments.
A customer who fails twice and succeeds on the third try is one successful order and three attempts. If your success rate uses attempts as the denominator, it looks like 33%; by orders it is 100%. Neither is wrong, but mixing them across weeks or providers produces false alarms.
How do the providers’ views compare?
| Need | Razorpay | PayU |
|---|---|---|
| Daily performance overview | Analytics Suite Overview with pinned KPIs (up to 4) and success-rate trends | Business Summary |
| Payment-level list | Payments and Orders views | Transactions |
| Settlement view | Settlements and reports | Settlements |
| Date comparison | “Compare with” another date range | Confirm in dashboard |
| Method filter | Filter by payment method | Confirm in dashboard |
| Gateway-level view | For Optimizer merchants, filter by gateway, including “No Gateway” abandons | Not applicable |
| Data history | From the time Analytics was enabled | Confirm retention in dashboard |
How do you diagnose a checkout failure?
Work through these steps with the same inputs in both dashboards.
1. Lock the frame
Set the date range, time zone and filters, and choose the denominator (orders or attempts). Compare the same weekdays to avoid weekend effects.
2. Compare periods
On Razorpay, use “Compare with” to set this week against last week. Look at success-rate trends by payment method. A fall concentrated in UPI points to different causes than one spread across cards.
3. Follow the flow
Razorpay’s Analytics shows how payments flow across the checkout and where they drop, with failures analysed by source, step and reason. Separate customer-side reasons (cancelled, wrong PIN, insufficient funds) from bank, network or gateway reasons. Excluding customer errors changes how you interpret the trend, not the actual outcomes; report both numbers.
4. Check gateway routing (if you use Optimizer)
Optimizer merchants can filter by gateway. Razorpay labels checkouts abandoned before a gateway was chosen as “No Gateway”, so those are not wrongly blamed on a provider. See our guide to payment orchestration for how routing affects success rates.
5. Do the same on PayU
Use Business Summary and Transactions to look at success rates and failed transactions for the same frame. Where you cannot find a drill-down by step or reason, or the denominator is unclear, note it as a question for your PayU account team rather than assuming it is missing.
Diagnosis outcome table
| Pattern | Likely cause | Next step |
|---|---|---|
| Drop limited to one UPI app or bank | Bank or app downtime | Check provider downtime notices; show alternatives at checkout |
| Drop in cards with authentication failures | 3DS or issuer issues | Review by issuer; contact provider |
| Rise in customer cancellations | Checkout UX change or price shock | Review recent site changes |
| Rise in “No Gateway” abandons (Optimizer) | Pre-payment drop-off | Investigate checkout page, not gateways |
How do you close the ledger with exports?
Reconciliation needs a single row per payment that carries everything finance needs, and a way to link it to the bank credit.
Razorpay Custom Reports
Custom Reports let you select and combine fields from multiple Razorpay entities, such as payments, orders, refunds and settlements, into one report. Razorpay suggests using them when you regularly combine two or more standard reports or need multi-entity analysis. Our references indicate you can keep up to 15 custom reports with up to 50 columns each, editable only by their creator, and download them as CSV, Excel or text.
PayU reports
PayU documents transaction, settlement, refund and monthly TDR (transaction discount rate) reports in CSV or XLSX, which can be scheduled daily, weekly or monthly to up to two recipient email addresses per scheduled report.
Field map for reconciliation
| Field | Why you need it | Razorpay | PayU |
|---|---|---|---|
| Payment ID | Unique key per attempt | Payments entity | Transaction report |
| Order ID | Links attempts to your order | Orders entity | Transaction report (merchant reference) |
| Status | Captured, failed, refunded | Payments entity | Transaction report |
| Gross amount | Sale value | Payments entity | Transaction report |
| Fee | Gateway charge | Payments or settlement fields | Settlement / TDR report |
| Tax on fee | GST on gateway fee | Payments or settlement fields | Settlement / TDR report |
| Refund amount and ID | Deductions from settlement | Refunds entity | Refund report |
| Settlement ID | Groups payments into one bank credit | Settlements entity | Settlement report |
| Bank reference (UTR) | Matches the bank statement | Settlements entity, where available | Settlement report, where available |
With Razorpay, these can come from one custom report. With PayU, expect to join the transaction, settlement and refund reports on payment ID.
Worked example: the ₹3.2 lakh gap
Assumptions: monthly captured sales ₹48,00,000; bank credits ₹44,80,000.
| Item | Amount |
|---|---|
| Captured sales | ₹48,00,000 |
| Less: gateway fees and GST | ₹1,13,280 |
| Less: refunds deducted from settlements | ₹1,40,000 |
| Less: last day’s sales settling next month | ₹66,720 |
| Expected bank credits | ₹44,80,000 |
Once fees, refunds and the cut-off are visible on the same rows, the “gap” is explained. Anything left over goes to an exception queue: unmatched payments, late refunds, chargebacks or adjustments.
Keep three numbers distinct in every report: authorisation-based GMV, captured revenue and cash received in the bank.
What should your month-end checklist include?
- Raw files kept: store the original exports, not just pivot tables.
- Report definitions written down: which fields, filters and date basis each report uses.
- Cut-offs agreed: time zone and the last capture time included in the month.
- Stable join keys: payment ID and settlement ID, not customer names or amounts.
- Duplicate checks: the same payment should not appear twice across files.
- Access controls: limit who can create and download reports; see dashboard access controls.
- Reviewed adjustments: every manual adjustment has a reason and an approver.
- BI hand-off: if you load files into a BI tool, document the file-based process; do not assume a live connector or automatic ledger posting.
Who owns which number?
Analytics problems drag on when nobody owns the definition. A simple ownership table avoids that:
| Metric | Owner | Source of truth | Review cadence |
|---|---|---|---|
| Order success rate | Product or growth | Razorpay Analytics / PayU dashboard, orders denominator | Daily |
| Attempt failure reasons | Payments or engineering | Failure analysis by source, step and reason | Daily during incidents, weekly otherwise |
| Captured revenue | Finance | Payments export, captured status | Weekly |
| Fees and GST on fees | Finance | Settlement or custom report | Monthly |
| Refunds and chargebacks | Finance and support | Refund report | Weekly |
| Bank credits | Finance | Bank statement matched to settlement IDs | Weekly and at month end |
How to run a weekly payment health review
A 30-minute weekly review keeps both problems small:
- Success rate by method: this week against last week, same filters and denominator.
- Top three failure reasons: what changed, and whether they are customer, bank or gateway issues.
- Open incidents: downtime events and their customer impact.
- Settlement check: captured sales against expected settlements, with exceptions listed.
- Refund and chargeback trend: volume and reasons.
- Actions: owner and date for each fix.
With Razorpay, steps 1 and 2 come from the Analytics Suite and steps 4 and 5 from a saved custom report. With PayU, use the dashboard views and the scheduled transaction, settlement and refund reports. Keeping the same agenda every week makes trends obvious and stops one bad day from being mistaken for a crisis. If you run multiple gateways, add a gateway comparison using Optimizer filters.
Why we recommend Razorpay for teams that diagnose and reconcile in one place
For payments and finance teams that need both checkout diagnosis and a defensible month-end close, we recommend Razorpay when its reporting fits the business. Our reasons:
- Diagnosis built in: date comparison, payment-method filters, payment flow and failure analysis in the Analytics Suite.
- Routing clarity for Optimizer merchants: gateway filters and a separate “No Gateway” bucket keep abandons from distorting gateway comparisons.
- Multi-entity exports: Custom Reports combine payments, orders, refunds and settlements, reducing manual joins at month end.
- Flexible formats: CSV, Excel or text downloads for spreadsheets and BI tools.
These are reporting-fit reasons. They are not a claim of higher success rates or measured time savings.
When PayU may fit better
PayU may suit teams whose main need is regular, predictable files delivered to finance: transaction, settlement, refund and monthly TDR reports in CSV or XLSX, scheduled daily, weekly or monthly. Existing PayU merchants with established spreadsheet processes may prefer to stay. Confirm which success-rate drill-downs and denominators PayU’s dashboard provides before relying on it for checkout diagnosis.
FAQs
1. How do I find out why my payment success rate dropped on Razorpay?
Use the Analytics Suite: compare the period with an earlier date range, filter by payment method, and review payment flow and failure reasons by source and step. Separate customer-side reasons from bank or gateway issues before deciding on action.
2. Should success rate be calculated on orders or attempts?
Either can be valid, but be consistent. Order-based success shows whether customers eventually paid; attempt-based success shows how often individual tries fail. Repeated attempts inflate failure counts if you mix the two.
3. What are Razorpay Custom Reports used for?
They combine selected fields from multiple Razorpay entities, such as payments, refunds and settlements, into one report. They help with reconciliation, tax and compliance work where you would otherwise merge several standard reports.
4. Can PayU email reports automatically?
Yes. PayU documents scheduling reports daily, weekly or monthly, with up to two recipient email addresses per scheduled report, in CSV or XLSX format.
5. Why don’t my bank credits match my sales report?
Usually because of fees and GST on fees, refunds deducted from settlements, chargebacks and cut-off timing between capture and settlement. Map each payment to its settlement ID and bank reference to find the difference.
6. Does Razorpay Analytics show historical data before it was enabled?
No. Razorpay notes that data is available from the time Analytics was enabled on the account, so plan baselines accordingly.
7. Do these reports post entries to my accounting software automatically?
The reports described here are exports. Moving them into accounting or BI tools is a file-based process unless you use a separate integration.
8. How long should we keep raw payment exports?
Keep them for as long as your accounting and audit policy requires, typically several years for financial records in India. Store original files unchanged, with report definitions, so any reconciliation can be reproduced later.