There is a genre of core banking content that has become almost liturgical. It opens with customer expectations — real-time, instant, seamless. It notes that fintechs enjoy a structural advantage, unburdened by legacy. It lays out the migration patterns: phased, hollowed-out, sidecar. It closes on governance, testing, and parallel run.

Visa's recent BrandVoice piece on core migration, built around its $1 billion acquisition of Pismo, follows the form closely and does it well. The taxonomy is correct. Sequencing payment-heavy products first to stabilize revenue-critical flows is what competent programs do. Hollowing out the core — demoting the legacy system to a thin ledger of record while services migrate outward — is a real and widely used pattern. Sidecar cores for new products and segments are real. Dual-run reconciliation is the right integrity control.

None of that is the problem. The problem is one anonymized noun.

Who ran the playbook

The piece cites Federal Reserve Bank of Kansas City research on phased, component-based migration, describing "one US-based national bank" that upgraded its lending system before its deposit system, ran thousands of simulations, operated old and new systems simultaneously, and migrated branches incrementally.

The underlying Kansas City Fed briefing names the bank. It is Zions Bancorporation — a regional operating in 11 states with roughly $90 billion in assets at the time. Zions' executive vice president described the strategy publicly at the 2022 Digital Banking Conference, including the reasoning for starting with lending: lower customer visibility meant conversion problems would do less damage.

That is a real and instructive case. It is also an institution with in-house technical capacity, dedicated program leadership, and the negotiating weight of a top-tier regional. The single named example of successful migration in a piece addressed to "traditional financial institutions" is roughly 180 times the size of the median US bank.

This is the gap the genre never closes. Every one of these strategies presupposes agency: that an institution can decide which architecture it wants and go execute against that decision. For Zions, that assumption holds. For most institutions by count, it does not — and the reason is not architectural. It is contractual.

The constraint

The Kansas City Fed's companion briefing on core provider market structure is unusually direct for a Fed publication. The Big Three — Fiserv, FIS, and Jack Henry — collectively served more than 70 percent of banks surveyed in 2022 and close to half of credit unions surveyed in 2020. Fiserv alone accounted for 42 percent of banks; Jack Henry 21 percent; FIS 9 percent.

Concentration alone would merit attention. What makes it decisive is the shape of the agreements underneath it. The Fed briefing catalogues the problems plainly:

• In ABA's 2020 survey, more than half of responding banks identified fees charged for third-party integration and fees charged for upgrades as the most problematic contract provisions. Both remained among the top three in the 2022 survey.
• Some contracts assign the core provider sole ownership of work related to core functioning — software, source code, modifications, enhancements.
• Ownership of the institution's own customer data is frequently ambiguous in contract terms. Where the provider owns it, switching costs rise accordingly.
• Deconversion fees and early-termination charges are often unclear. The Fed cites characterizations of these contracts as complex and tome-like, running to hundreds of pages with duration terms that are difficult for less sophisticated clients to parse.
• Some banks have litigated. Millington Bank sued FIS in 2020 over what it characterized as coercive contract tactics.

The lock-in shows in the tenure data. In ABA's 2022 survey, 61 percent of banks had been with their core provider for more than a decade, and another 20 percent for five to ten years.

ABA's most recent round, fielded in 2024 and released in February 2025, adds the detail that matters most. Overall satisfaction sits at 3.19 on a five-point scale — up from 3.01, still poor. Just 53 percent of bankers are satisfied; 35 percent are dissatisfied. Only one in five expects to switch. And satisfaction declines steadily across the contract term, reaching its lowest point precisely as renewal approaches.

That last finding is the whole story in one line. Dissatisfaction peaks exactly when the institution has the least room to act on it.

The strategy that requires your incumbent's permission

Of the migration patterns, hollowing out the core sounds most accessible to a smaller institution. It is incremental. No big-bang cutover. Keep the ledger, move capability outward.

It also requires something the playbook never states plainly: your existing core provider has to let you do it.

Hollowing out depends on moving data and transactions in and out of the legacy ledger through interfaces you control. Recall what banks told the ABA were their two most problematic contract provisions — fees for integrating third parties, and fees for upgrades. Those are not incidental line items. They are precisely the levers that determine whether this pattern is affordable. An institution facing per-integration charges, or a bundling arrangement requiring it to buy the provider's own version of the service, is not choosing between architectures. The pattern is priced out of reach.

Zions negotiates past this. A $600 million community bank on largely standardized paper does not. The strategy that looks most incremental and least risky is, in practice, the one most dependent on negotiating power the smallest institutions lack.

Sidecar cores carry a related mismatch. Standing up a greenfield platform alongside the incumbent answers a growth problem — new segments, products, markets. Most community institutions are solving a cost problem, and a sidecar adds a second platform to run, staff, and reconcile while the expensive one stays put.

There is also a question of what is actually being sold. Visa acquired Pismo as a cloud-native issuer processing and core banking platform, and market analyses of modern card issuing place it in the competitive set with Marqeta, Galileo, i2c, and CoreCard. Its reference clients point the same way: Citi, Itaú, Revolut, N26, Nubank — card and program businesses, not deposit franchises. Pismo has genuine core banking capability, and this is a matter of emphasis rather than mislabeling. But its center of gravity is issuer processing, while the article's frame is deposit-side modernization: the ledger of record, the system a community bank replaces when it leaves a legacy platform. Those are adjacent markets, not the same one.

What failure actually looks like

The Visa piece contains no failures. Every example succeeds. For a piece about operational continuity and downtime, that absence is the tell.

TSB is the case the genre avoids. In April 2018, under Sabadell ownership, TSB migrated onto Proteo4UK, a platform built and tested by SABIS, Sabadell's IT division. All of TSB's branches and a significant share of its 5.2 million customers were affected. The bank did not return to business as usual until December 10, 2018. It paid £32.7 million in customer redress, and in December 2022 the FCA and PRA fined it £48.65 million — £29.75 million and £18.9 million respectively, after a 30 percent settlement discount. TSB's former CIO was personally fined £81,620.

Now the detail that should reframe how the Visa piece reads: the regulators found that the data migrated successfully. The platform failed immediately anyway, on configuration, coding, and capacity.

Visa devotes its longest section to data integrity — cleansing, mapping, reconciliation, post-migration audit. All sensible. None of it would have prevented the most consequential core migration failure of the last decade. The regulators' findings centered on planning, testing scope, governance escalation, and — most relevant to a piece selling a vendor platform — failure to manage operational risk arising from outsourcing to a critical third-party supplier.

An article about migration risk that omits the vendor-dependency failure mode, published by a vendor, is not describing risk management.

What is actually moving

It would be dishonest to present this as static. Pressure has produced real change, and the change came from exactly the places the architecture conversation ignores.

In 2020, FIS introduced a simplified contracting model for qualifying community banks and credit unions that removed required term lengths, liquidated damages, and exclusivity requirements, and defined deconversion fees explicitly. Jack Henry announced plans to unbundle its services onto the public cloud and now supports integration with more than 200 vendors. CFPB Director Rohit Chopra publicly raised the power imbalance between smaller institutions and the Big Three in 2022. A coalition of banking and legal executives — the Golden Contract Coalition — formed in 2016 specifically to negotiate collectively against core providers.

Two data points deserve particular weight. Seattle Bank, a single-branch institution with $650 million in assets, fully replaced its Fiserv core with Finastra — a decision driven by dissatisfaction with inflexibility and fee structure. Small institutions can do this. And in ABA's 2024 survey, the highest-rated core provider was COCC, which is member-owned by the institutions it serves.

That last fact is not a coincidence, and it points somewhere the architecture debate does not: when the institutions own the vendor, the contract problem largely dissolves.

The counter-pressure is real too. Finxact, the most prominent next-generation challenger, was acquired by Fiserv in 2022. Consolidation absorbs the alternatives that competition depends on.

The unanswered question

If the constraint is contractual, the interventions that matter are contractual.

Contract transparency — term length, renewal mechanics, notice windows, deconversion fees, integration consent rights — determines who can modernize far more than any architecture diagram. Data portability determines what leaving costs. And core providers are examinable under the Bank Service Company Act; whether that authority is exercised in proportion to a market where three firms serve over 70 percent of banks is a legitimate open question, and one worth reporting out.

Core modernization is real, necessary, and achievable. The patterns Visa describes are the right patterns. But a playbook whose sole named exemplar is a $90 billion regional, published by a vendor selling into the modernization, and silent on the contractual machinery that determines who can run it, is describing a decision most of its nominal audience cannot currently make.

The architecture question is solved. The market structure question is not.

Sources

Federal Reserve Bank of Kansas City, "Core Banking Systems and Options for Modernization," Payments System Research Briefing, February 28, 2024 — Zions case detail, three modernization options, Seattle Bank, Mascoma Bank, data ownership ambiguity. https://www.kansascityfed.org/research/payments-system-research-briefings/core-banking-systems-and-options-for-modernization/

Federal Reserve Bank of Kansas City, "Market Structure of Core Banking Services Providers," Payments System Research Briefing, March 27, 2024 — concentration figures, contract provisions, ABA survey citations, Chopra remarks, Golden Contract Coalition, FIS contracting model, Millington Bank litigation. https://www.kansascityfed.org/research/payments-system-research-briefings/market-structure-of-core-banking-services-providers/

FCA / Bank of England, "TSB fined £48m for operational resilience failings," December 20, 2022 — penalty split, customer impact, timeline, outsourcing findings. https://www.fca.org.uk/news/press-releases/tsb-fined-48m-operational-resilience-failings https://www.bankofengland.co.uk/news/2022/december/tsb-fined-for-operational-resilience-failings

American Bankers Association, "ABA Survey: 53% of Banks are Satisfied with Their Core Platform Provider," February 2025 (2024 survey round) — satisfaction scores, switching intent, satisfaction decline across contract term. https://www.aba.com/about-us/press-room/press-releases/core-platform-survey-2025

Visa, "Visa Completes Acquisition of Pismo," January 16, 2024 — transaction detail. https://usa.visa.com/about-visa/newsroom/press-releases.releaseId.20301.html