How Data Catches Silent Churn Before Cardholders Walk Away

Watch more: The Digital Shift: Visa DPS, Nick Roberts 

Data has become the tightest lever in card issuing. It protects portfolios, fuels growth and keeps operations lean. But only if banks can turn raw transactions into decisions at the speed of a tap. That is the bet behind Visa DPS’ data strategy and, in particular, its Data Manager platform. It aims to make high quality card and account data both trustworthy and immediately useful to the people who run an issuing business.

“Issuers want data that includes a lot of profitability metrics [and] fraud indicators,” Nick Roberts, senior director for Visa DPS product management in data and analytics, said in a PYMNTS interview. “It allows the issuer to identify those high‑value cardholders and really optimize spend.”

Issuers, Roberts said, want less friction in how they get data. They don’t want another dashboard that sits apart from day‑to‑day work. Data Manager starts with an enriched dataset that brings disparate DPS tools into one application, then lets clients consume that information in the way that fits their organizations: a web interface for business users, APIs and cloud integrations for developers, and compatibility with in‑house business‑intelligence tools for analysts.

That flexibility extends to who uses the data and when. Executives can log in for roll‑up KPIs and portfolio views, while data teams can pull premium extract files into their own warehouses for modeling and experimentation. Business analysts, meanwhile, can work from near‑real‑time transaction detail in the same environment where they see executive‑level reporting. The design is meant to close the gap between analysis and action. 

Roberts described a delivery model organized around user roles rather than one‑size‑fits‑all reports. Although availability may vary depending on the receiving financial institution, region or other external factors, Data Manager offers near-real-time access to transaction-level data for investigators and operators, executive dashboards for leaders, and bulk extracts for data science teams that prefer to work within their own tooling. The aim: let each group get what it needs without tickets or handoffs.

“We’re just seeing more need for flexibility,” Roberts said, “and we want to be able to deliver [data] in the desired way of the use case.” 

Much of the operational return comes from frequency of use. Roberts said most users access Data Manager several times a week, if not daily, to self‑serve answers that would otherwise require manual queries or custom reports. Eliminating that lag matters when conditions change quickly.

In one recent scenario, issuers needed to know which cardholders lived in (or had recently spent money in) areas hit by natural disasters. Because the relevant data was already normalized and searchable in Data Manager, clients could compile targeted contact lists, share nearby ATM locations and push emergency support information without waiting for a batch file. 

That same principle applies to revenue protection. If a high‑spend cardholder’s monthly transaction count falls from 20 purchases to seven, Roberts said, issuers can spot the drop and intervene before attrition sets in. The idea is to catch movement at the edge and trigger outreach, offers or service fixes while there’s still time to change the outcome. 

“AI [artificial intelligence] is changing what we’re doing with data and analytics,” Roberts said. Instead of analysts spending hours aggregating data before they can even interpret it, he expects agentic tools to assemble the relevant datasets and return answers in plain language. In his view, that turns the analyst from a data gatherer into a decision maker. This can make insights both more accessible and more timely across the issuing organization. 

That shift also changes the timing. Rather than relying on historical reviews to decide what to fix next month, Roberts said Visa DPS is building toward alerts that fire within hours, even minutes, of a measurable change. Artificial intelligence will “identify these trends, flag anomalies and suggest next steps,” and in some cases execute actions automatically. The outcome is a portfolio that moves from lagging indicators to early signals, and from reactive reporting to continuous course correction. 

Roberts ties that speed directly to P&L. Data Manager bakes in profitability metrics, fraud indicators and merchant level reporting, along with issuer defined roll-ups, to surface where returns are strongest and where spend is at risk. Visa DPS is also rolling out a Goal Tracker app that lets issuers set performance targets, benchmark against peers and monitor progress in one place. AI will soon recommend goals based on current performance and peer context. Ultimately, Roberts expects an agent to accept a goal and either propose the plan or carry out defined actions to hit it. 

Risk and marketing teams are part of the loop, too. Because Data Manager includes fraud rule performance and transaction context, issuers can tune controls based on evidence rather than guesswork. On the growth side, cardholder segmentation, demographics and integrated connections to Visa Campaign Solutions allow teams to target, launch and measure campaigns inside a closed loop from audience to outcome without exporting sensitive data to multiple systems. 

Roberts’ roadmap points to deeper integrations across Visa tools and more automation atop trusted data. “Our goal is to make data not just accessible, but predictable,” he said, “and then enable issuers to “make automated actions on top of it so issuers can respond to market changes almost instantly.”

The destination is clear: decisions arriving as quickly as the transactions that trigger them, with less effort and better results. 

The post How Data Catches Silent Churn Before Cardholders Walk Away appeared first on PYMNTS.com.