PAYMENTS FM
PAYMENTS FM
Payments Data for Business Success: Albert Drouart, Pagos
0:00
-1:29:20

Payments Data for Business Success: Albert Drouart, Pagos

Albert Drouart from Pagos joined PAYMENTS FM to talk about payments data, monitoring, benchmarks, AI, and how merchants can use payments data in day to day work.

Note: We are launching the State of Payments survey. It takes about four minutes to complete, and anonymous responses are welcome. We will publish the results at the end of the year.

Why this matters

Payment teams usually know when a provider is down. The harder question is what happens when the provider is up, checkout is live, and payment performance still hurts the business.

Good customers can get declined. Fraud rules can block too much volume. Retries can add cost. Chargebacks can move in one market. Fees can change in ways finance needs to explain.

This is why payments data needs a business context. Networks change. Issuers change. Customer behavior changes. Internal teams ship product changes. Revenue can leak quietly if nobody reviews the data with the right context.

What to watch

Payment teams should watch the signals that connect payments to revenue, cost, customer experience, and risk.

  • Approval rate by payment method and other parameters

  • Technical declines by provider, integration, and setup

  • Decline codes that affect retries and customer messaging

  • Fraud rules that block good customers

  • Chargebacks by segment and market

  • Payment fees and avoidable retry cost

  • Retry performance and recovery rate

  • Subscription payment failures after internal changes

  • Wallets, BNPL, ACH, and pay by bank performance

  • Market changes that affect authorization rates

What it means for your team

Payments data is useful for many teams.

Product needs it for checkout conversion and customer experience. Engineering needs it for integrations and releases. Finance needs it for cost and reconciliation. Risk and fraud teams need it for exposure and false positives. Operations needs it for incidents. Leadership needs it for revenue.

Albert explains why merchants need clean data, useful alerts, benchmarks, and enough context to decide what to do next.

This also applies to single PSP merchants. One provider still leaves many questions: approvals, declines, cost, chargebacks, fraud rules, issuers, payment methods, markets, and customer behavior.

What to do next

Start with the places where payment performance can quietly leak money.

  • Review technical declines every week

  • Compare approval rate by payment method and other parameters

  • Connect payments to customer, order, subscription, and revenue data

  • Check if retries recover revenue or add cost

  • Review fraud rules for good customers blocked

  • Track chargebacks and fraud by segment

  • Monitor payment fees with finance

  • Give product, finance, risk, fraud, operations, and engineering one shared view

  • Use benchmarks before assuming the issue is internal

  • Clean PSP data before adding more dashboards or AI workflows

Questions to ask internally

  • Where do we see payment performance drop before revenue notices it?

  • Can we tell if a decline issue comes from us, the issuer, or the market?

  • Do product, finance, risk, fraud, operations, and engineering use the same payment data?

  • Which retries recover money, and which retries add cost?

  • Which declines can we actually fix?

  • Which fraud rules block good customers?

  • Can finance explain payment cost changes quickly?

  • Do benchmarks show if the issue is internal or market wide?

  • Is the data clean enough for useful AI work?

Guest perspective

  1. Payment monitoring should cover more than provider uptime. A provider can be online while approval rates, fees, fraud rules, or chargebacks move in the wrong direction.

  2. Raw PSP data is useful, but merchants still need cleanup. The work often includes normalized fields, business context, duplicate removal, and links to customer or order data.

  3. Benchmarks help teams avoid guessing. If approval rates fall, the issue may come from an internal change, issuer behavior, card network changes, or a broader market pattern.

  4. AI is useful when the data is ready. Machine learning can help with detection and anomaly analysis. Generative AI can help summarize changes and explain likely causes.

  5. The real work is turning payments data into operating information that teams can use.

Listen or watch

Resources

Related episodes

Subscribe

PAYMENTS FM helps merchants, platforms, marketplaces, and payment product teams understand how payments work in practice through podcast episodes, expert interviews, community conversations, and practical payments education. Subscribe for practical payments education and operator-focused conversations about money movement.

Discussion about this episode

User's avatar

Ready for more?