How to spot revenue leakage you didn't know you had
Industry
Updated 4 Aug 2026
5 min
Bozhena Labinska
Merchant Integrations Lead, Solidgate
Bozhena Labinska, Merchant Integrations Lead at Solidgate, and Tom Chmielewski, VP of EMEA & APAC at Testlio, recently sat down for a joint webinar on the payment failures that never trigger an alert. Here's what to check before your next payment method launch.
I recently joined Tom Chmielewski, VP of EMEA & APAC at Testlio, for a webinar on the payment failures that never make it into an incident report.
Solidgate has been running live payment tests with Testlio since 2025, covering 150+ countries and 800+ payment methods. Solidgate sees what's happening inside the payment ecosystem. Testlio brings the other half of the picture by showing what real customers actually experience. When you combine those two perspectives, you start finding revenue leakage issues that neither side could fully explain on its own.
Our work keeps surfacing the same lesson for merchants: the most expensive are usually the quiet ones. They're not outages. They don't trigger alerts. They just quietly stop converting, and a healthy dashboard won't tell you why.
Here's what that means in practice.
Revenue leaks before a payment ever gets declined
A merchant launches a new . The integration is correct. Transactions process as expected, and every dashboard reads healthy. Then some customers get redirected to their banking app to authorize the payment and never make it back to checkout.
The works exactly as designed. The customer journey doesn't. Payment data tells you customers aren't completing the purchase, but what it doesn't tell you is why. This may be a browser issue, a confusing redirect, or a flow that just feels unfamiliar the first time someone pays this way.
One example stands out. A merchant's integration looked perfectly healthy – transactions created, flow tested, every dashboard green. But when someone went through the entire purchase journey as a real customer, they found something the dashboard missed. The customer got access to the product before the payment was actually confirmed.
Nothing had technically failed – the business process just didn't match the payment lifecycle. Some of those payments were later declined or reversed, after the customer already had the product.
That kind of revenue leak usually shows up as a small percentage of orders, every month, quietly written off as bad luck. But add that up over a year, and it's real, ongoing lost revenue.
A working integration doesn't always mean a working payment experience
Ask ten people in payments what a successful integration means, and most describe the same thing: the API call went through. At Solidgate, that's not the bar we hold ourselves to.
Take in Brazil. The technical integration can pass every test – endpoints respond correctly, statuses update as expected – and still fail customers if the flow doesn't match how Brazilian consumers expect to pay. Every local payment method carries its own expectations, its own flow, and its own behavior.
That plays out the same way across every market we've helped merchants expand into – Bizum in Spain, UPI in India, GCash in the Philippines. Integration works, transactions process, dashboards say it's fine. But adoption is still lower than expected.
Investigate, and the reasons are usually simple:
- The payment method isn't shown to the right customers – a French Open Banking option shown to German customers
- It's buried too low in the instead of sitting among the primary options
- The flow follows the technical spec but doesn't match what local customers expect
There's no error to chase, no ticket to file. Customers just don't use it.
What to test for, then, is whether customers recognize the payment method, trust it, and come back after in their banking app. Those moments influence payment success as much as the technology behind it.
That's exactly the mindset we applied during our Pix Automatic launch, working closely with Testlio to test real customer journeys before merchants encountered any edge cases in production.
Crowdsourced QA catches what internal teams miss
Internal teams test what they know to test for. The blind spot is exactly what they don't know to check – because they're not the customer in that market, on that device, using that payment method for the first time.
That's the reason Solidgate works with Testlio's crowdsourced testers. Real users, in the actual markets a payment method serves, complete the purchase the way their peers would – without a script, and without knowing in advance how the flow is supposed to work.
It's how we caught the case above: a dashboard that looked fine couldn't show that access was granted before payment confirmation. A real customer walking the journey did.
We used the same approach ahead of our Pix Automatic launch, running end-to-end validation with Testlio to test real customer journeys before merchants encountered any edge cases in production. It's rarely a broken API that crowdsourced testing catches. It's a customer journey that doesn't behave the way an internal team assumed it would.
No need to check everything – start with the outlier
Don't try to look at everything at once. Start with the area you're least confident about – a new market, a payment method launched a few months back, or a segment you don't have much visibility into.
From there, use data to narrow the search. Compare approval rates, conversion, and drop-off across markets, payment methods, devices, and browsers. The outliers tell the story, not the average. Once you find one, don't assume you know the cause – walk the journey yourself, or ask someone who actually matches your target customer to complete the purchase exactly as they normally would.
This is why we always recommend User Acceptance Testing before launch and after significant changes – and it's where working with a partner like Testlio makes a real difference. They validate the payment experience with real users, in real markets, under conditions that are almost impossible to recreate internally.
And when a failure sits at the boundary between your platform and your own implementation, resist the urge to assign blame before you've diagnosed anything. Define ownership by layer and investigate jointly – the problem doesn't care who's responsible, and neither side usually has the full picture alone.
Do this consistently, and you stop chasing a vague feeling that something's wrong. You're looking at one specific journey, for one specific segment, with a real person walking through it in real time – a much faster way to find a leak than waiting for next quarter's report to confirm it.
Payment quality was never really about payments. It's about whether the experience matches what you designed it to be.
Want the full conversation? with Tom Chmielewski and Testlio.
If your current setup can't tell you whether revenue leakage is costing you right now, .