At the 2026 Merchant Advisory Group (MAG) conference in San Diego, a session titled From Data to Decisions: How AI is Transforming Merchant Payment Insights explored a challenge many payments teams face: making sense of fragmented transaction data.
One company shared how they tackled it.
Moderator Kiel Cook was joined by Tanit Parada Tur, Payments Operations Manager at Grover, and Marco Conte, VP of Data Operations at IXOPAY, to discuss how Grover turned payment data chaos into actionable insights.
With hundreds of thousands of transactions per month and a recurring subscription model, Grover needed a way to unify scattered payment data. Enabling it to uncover the insights needed to improve performance and reduce costly payment failures.
The Core Problem: Fragmented Data and Conflicting KPIs
If you work in payments, the scenario Tanit described may sound familiar. You ask a simple question about performance and end up pulling reports from three different tools before anyone can even agree on the numbers.
As she explained during the session, “there was a lot of fragmentation when it comes to data.”
At Grover, payment data lived in multiple places:
PSP dashboards
Tableau reports
Internal spreadsheets
This fragmentation created a bigger operational problem. When teams asked how a payment method was performing, they often received multiple conflicting answers depending on who ran the report.
There was more than enough data. The real issue was the lack of a shared view of performance.
For payments teams, this creates several challenges:
Decisions take longer because data must be reconciled across tools
Finance, product, and payments teams interpret KPIs differently
Optimization becomes reactive because issues are discovered late
Recurring models add complexity around retry logic and decline analysis
For many at the Merchant Advisory Group conference, this challenge resonated immediately. As transaction volumes grow and payment stacks expand, fragmented payment data can turn everyday operational questions into time-consuming investigations.
Where Was Grover Leaking Money?
According to Tanit, once the payments team at Grover began digging into the data, two critical questions emerged:
Where are we leaking money?
Why are transactions failing?
Without a unified view of payment performance, answering these questions was difficult. The team needed visibility across key cost and risk drivers, such as:
Chargebacks
Scheme and PSP fees
Payment declines
As Tanit explained during the session, when you operate with multiple PSPs, it can be difficult to identify the root cause of an issue. Because there is no single view of performance.
With the help of IXOPAY’s solution, Grover uncovered a key problem. A significant number of declines were occurring for an unexpected reason. An integration error was causing the system to retry hard declines, triggering steep network penalties and quietly increasing costs.
Fixing the integration immediately reduced unnecessary retries, lowered penalties, and improved transaction efficiency.
The insight itself was simple. What changed everything was finally being able to pinpoint it amidst the noise of fragmented payment data.
From Spreadsheets to Payment Intelligence
As Marco Conte explained, payment operations become significantly more complex as businesses scale. Higher transaction volumes introduce multiple layers of analysis across gateways, billing models, and retry logic.
Payments teams often deal with factors such as:
Multiple payment gateways
Recurring billing cycles
Retry logic for failed transactions
End-of-month penalties from payment networks
Traditionally, investigating performance meant:
Exporting raw data
Building massive tables across many dimensions
Manually analyzing hundreds of transaction combinations
AI changes how teams approach that analysis. It helps surface patterns that would otherwise take hours to uncover, such as:
The largest drops in authorization rates
High-value anomalies across gateways or banks
Fee and cost patterns
Early warning signals for potential penalties
Instead of discovering problems at the end of the month, teams like Grover can now identify risk within the first few days of a billing cycle.
Anomaly Detection: The Breakthrough Feature
Among the many AI capabilities discussed during the session, Tanit highlighted anomaly detection as the most powerful feature Grover currently uses.
Payment environments generate too many variables for teams to anticipate every issue through manual alerts. AI anomaly detection helps surface patterns that would otherwise have remained buried in large datasets, including:
Unusual combinations of a specific PSP and issuing bank
For example, authorization rates might drop only when transactions are routed through a particular PSP and processed by a specific issuing bank in one region. The pattern may be invisible in overall approval metrics, but becomes clear when the data is analyzed across multiple dimensions.Low-volume anomalies that could grow into larger problems
A small cluster of declines from a newly added payment method or market may appear insignificant at first. However, anomaly detection can flag the trend early. Giving the team a chance to investigate before it begins affecting larger transaction volumes.Performance patterns that are difficult to detect through manual analysis
Subtle shifts in authorization rates, retry outcomes, or fee patterns across hundreds of gateway, country, and payment-method combinations can be hard to spot in spreadsheets. AI can scan these combinations, surface where performance changed most significantly.
These signals act as prompts for investigation. The system flags the anomaly, but the payments team determines the next step.
At Grover, the team uses these signals to:
Investigate potential issues
Prioritize which anomalies require action
Apply business context to interpret the data
Maintain full ownership of operational decisions
The approach reflects a key theme from the session. AI accelerates analysis, while human expertise drives the final decision.
The Foundation: Clean, Unified Data
Marco emphasized a key point that often gets overlooked when merchants talk about AI. The technology can only deliver insights if the underlying data is collected, cleaned, and structured properly.
In modern payment stacks, data often arrives from many sources:
Multiple payment gateways
PSP reports and settlement files
Internal transaction logs and billing systems
Each system may format data differently. For example, the same decline reason or fee type can appear under different labels depending on the gateway. Without normalization, comparing performance across providers becomes difficult.
For merchants operating with ten or more gateways, this challenge can grow quickly. Data must be gathered, standardized, and transformed before it can support meaningful analysis.
Results and Business Impact
By consolidating data and applying AI-driven analysis with IXOPAY, Grover gained a clearer and faster way to understand payment performance. The payments team established a unified KPI framework. This helped finance, product, and operations teams work from the same set of metrics.
Greater visibility led to several tangible improvements:
Reduced penalties after identifying and correcting hard-decline retry behavior
Clearer insight into decline reasons across multiple PSPs
Faster investigation cycles when performance shifts occur
Earlier detection of potential payment risks
Stronger cross-functional alignment around payment KPIs
Greater confidence when making optimization decisions
For payments teams used to digging through exports and spreadsheets, the change was significant. Questions that once required hours of analysis could now be answered in seconds.
A Blueprint for Merchants
For the audience at the MAG conference, the takeaway was clear. Payments teams are not short on data. Rather, they’re often overwhelmed by fragmented and unstructured data.
Grover’s experience offers a practical framework that other merchants can follow:
Centralize and normalize payment data across gateways and providers
Establish shared KPIs across payments, finance, and product teams
Use AI to surface insights while keeping human expertise in control
Move from reactive reporting toward predictive performance management
With the right foundation in place, AI becomes a powerful tool for uncovering issues earlier and optimizing payment performance faster.
Want to explore how this works in practice? Take a closer look at IXOPAY’s payment intelligence capabilities, including anomaly detection.