ISO 20022 Puts Data Quality at the Centre of Cross-Border Payments
Cross-border payments have passed the headline milestone in the move to ISO 20022. On paper, the industry has done the hard part: messages have been mapped, formats changed and migration programmes closed. In practice, the competitive and supervisory story has moved on. What separates payments providers now is not the syntax of a message but the quality, completeness and consistency of the data it carries.
That distinction matters because ISO 20022 is not simply a replacement message format. It is a richer data model that can carry structured addresses, purpose codes, legal entity identifiers and more detailed remittance information. A payment can be technically valid and still fail to process cleanly because a purpose code is missing, an address is unstructured, a remittance field is truncated or a name and account do not match. The remaining work is therefore semantic rather than syntactic.
The remaining work is semantic
Many institutions treated ISO 20022 as an IT change programme. Field mapping was owned by technology teams, testing focused on message validation, and sign-off was tied to migration completion. That approach can satisfy a test pack while leaving the operating model unchanged. It does not ensure that the data entering the payment chain is complete, that code lists are used consistently, or that front-office and operations teams understand why a missing purpose code matters.
Semantic quality is determined upstream, at client onboarding, in payment initiation channels, in static data and in the controls that reject incomplete instructions before they enter the network. Structured postal addresses are a useful example. A free-text address may pass through a legacy system, but it is harder to validate, screen and repair. A structured address with defined elements is more useful to every party in the chain, provided those elements are populated consistently and not simply copied into the first available field.
Purpose codes present a similar challenge. The code list is only effective if institutions use it with discipline. A generic or inaccurate code may satisfy a validation rule, but it weakens analytics, screening and reconciliation. The same is true of remittance information. Richer data can support automation, but only when it is complete enough to identify an invoice, a customer or a transaction purpose without manual interpretation.
Data gaps propagate through correspondent chains
Cross-border payments are not a single hop. A payment may pass through several correspondent banks, each with its own processing rules, cut-off times and screening systems. If an intermediary truncates or reformats remittance information, the beneficiary may receive funds without the data needed to reconcile them. That is not a theoretical inconvenience. It creates manual repair, duplicate queries and breaks in cash application, particularly for corporates managing high volumes across multiple accounts and currencies.
Screening is another pressure point. Sanctions and anti-money-laundering controls rely on names, addresses, identifiers and transaction context. Poor data increases false positives, and false positives consume analyst time and delay payments. Richer, consistent data can improve matching, but only if the data is reliable across the chain. A single weak link can force the receiving institution back to manual review, undermining the value of better data elsewhere.
This is why data quality in payments is not a bilateral issue between a bank and its client. It is a network issue. Institutions need to think about how their data survives the journey, not just how it leaves their own systems.
Services built on data need comparable quality
Pre-validation, payment tracking and exception management are among the most tangible benefits of richer payment data. Pre-validation can catch formatting and data errors before a payment is sent. Tracking can show where a payment is and why it is held. Exception management can route cases with context rather than forcing operations teams to reconstruct the payment from fragmented messages.
These services are only as good as the weakest participant in the chain. A bank cannot pre-validate data it never receives, and a corporate cannot track a payment if status messages are not standardised. Domestic instant payment schemes that already use ISO 20022 raise expectations further. Corporates that receive rich remittance data on domestic rails will ask why cross-border payments cannot match. The answer will depend less on the standard than on the discipline of the institutions using it.
The commercial upside is significant. Enriched payment data can support treasury, cash forecasting, working capital and reconciliation services. Banks that can demonstrate reliable data quality can build propositions on cleaner flows rather than treating payments as a utility. That makes data quality a revenue question as much as a compliance one.
Ownership, measurement and supervision
Accountability remains unsettled. Does payments data quality sit with operations, the data office, product teams or compliance? In many institutions it sits everywhere and nowhere. Governance should span operations, compliance and technology, with clear ownership of data fields, controls at onboarding and payment initiation, service-level expectations with correspondents, and metrics that reflect operational outcomes.
Migration completion is a poor metric. Straight-through processing rates, repair rates, exception volumes, screening false-positive rates and reconciliation breaks are better indicators of whether the data programme is working. Supervisors are increasingly attentive to operational resilience and financial crime controls, and poor payment data touches both. Institutions that cannot evidence reliable data quality may find themselves explaining manual workarounds rather than demonstrating control.
The ISO 20022 migration was necessary but not sufficient. The next phase is about discipline: consistent codes, complete addresses, preserved remittance information and clear ownership across the payment chain. Institutions that can demonstrate reliable data quality will differentiate themselves in transaction banking. Those that cannot will absorb rising manual repair costs and supervisory attention. The standard has changed the message. The competitive advantage now lies in the data.