Platform capability
Provider performance, side by side
See how each provider actually performs on your traffic, by method, bank and hour, and set routing on evidence instead of habit.
- Measured on your own payments
- Cut by method, bank, amount and hour
- Declines split by who caused them
A dashboard comparing UPI success rate, 95th percentile latency and a 24-hour trend for three providers. Provider B's success rate drops and its latency rises in the second half of the day.
The numbers
What's measured, and exactly how
A success rate means little unless you know what's counted. These are the definitions behind every number.
- Success rate
- Authorized payments divided by attempts that reached the provider. A customer who abandons before an attempt isn't counted, so a provider isn't blamed for a closed tab.
- Recovered payments
- Payments that failed on one provider and succeeded on a retry. They're credited to the provider that approved them, and the first provider records a failure.
- Response time (p50, p95)
- Time from Peneu's request to the provider's response. p95 matters more than the average, because the slow tail is where timeouts start.
- Pending ratio
- The share of UPI and netbanking payments with no final status after a set time. It's often the first sign of a provider or bank problem.
- Decline mix
- Failures grouped by cause: customer, issuer, provider or risk. Only the provider share is something routing can fix.
- Effective cost
- Fees per successful payment, based on the pricing you enter, including charges for failed attempts where your agreement has them.
Views
Cut the data the way problems show up
| View | The question it answers |
|---|---|
| Provider × method | Which PSP is best for UPI this week? |
| Issuing bank or netbanking bank | Is one bank failing everywhere, or only on one provider? |
| Hour of day | Are we losing payments at peak? |
| Amount band | Do high-value cards decline more on the aggregator than on the acquirer? |
| Decline category | Are failures customer-side (nothing to fix) or provider-side (route around them)? |
How teams use it
From a number to a routing change
- 01
Spot it
UPI success on Provider B fell six points after 2 pm while A and C held steady.
- 02
Check it's the provider
The decline mix shows provider timeouts rising, not customer declines, and it isn't limited to one bank.
- 03
Adjust routing
Lower B's weight for UPI, or add a rule for the affected bank. If B is failing outright, failover will already have paused it.
- 04
Confirm it worked
Compare the next few hours on the same traffic. Change one thing at a time so you know what moved the number.
Worth knowing
Compare like with like
A provider that only receives your hardest traffic, such as international cards or high-value tickets, will look worse than it is. Compare providers within the same method, bank and amount band.
When you're evaluating a new provider, give it a small random share of the same traffic before drawing conclusions. A few days of a 5–10% split tells you more than any sales deck.
Who uses it
Different teams, different questions
Payments and product
Which provider should carry each method, and when to change it.
Finance
What each provider really costs per successful payment, including failed-attempt charges.
Engineering and on-call
Whether a drop is a provider, a bank or a bug in your own checkout.
Evaluating it
Questions to ask about this, of any provider
Use these with any platform, including Peneu. For Peneu, what's available for your business is confirmed during onboarding.
- Are success rates broken down by method and failure reason?
- A drop caused by customers' insufficient balance isn't a provider problem.
- Over what period are figures calculated?
- Short windows are noisy; long ones hide incidents.
- Is cost shown per successful payment?
- The cheapest rate per attempt can be the most expensive per success.
- Can I export the data?
- You'll want to check it against your own records and use it in provider negotiations.
FAQ
Questions about provider performance optimization
Where does the data come from?
From payments processed through Peneu. Every attempt, response and timing is recorded as it happens, and provider dashboards aren't used, so every provider is measured the same way.
Can I compare against a provider I don't use yet?
Only providers carrying your traffic can be measured. A small split is the practical way to evaluate a new one.
Can I export the data?
Yes. Views can be exported as CSV for your own analysis or finance reporting.
Does this replace provider reporting?
For comparing providers, yes, because each provider reports on itself differently. For settlement and fee detail, provider reports still matter, and Peneu reconciles against them.
Works together with
Measure your providers properly
Tell us which providers you use today. We'll show you what a like-for-like comparison would look like on your traffic.
Last reviewed . Samples on this page are illustrative.