1. Business Role
Leading a participant-based business
Kelly led a company whose product depended on other institutions.
Banks supplied accounts and credit. Merchant-side institutions connected sellers. Technology partners helped deliver payment experiences.
Visa coordinated network activity and sold related capabilities. It was not one integrated consumer bank making every loan.
His operating problem was therefore different from choosing the features of a Visa-owned banking app.
The company needed banks, merchants and partners to keep choosing its infrastructure while commerce moved into new interfaces.
CEO and board responsibilities
Kelly later combined the CEO and board-chair roles.
After Ryan McInerney became CEO, Kelly remained executive chairman before retiring from the board.
A product announced during that transition belongs to a particular company period; it should not automatically be described as Kelly's personal initiative.
2. The Business He Inherited
A network that was already international and digital
Visa had an established processing network, institutional relationships and digital-payment capabilities before Kelly became CEO.
He did not start with a domestic paper-card business waiting for its first electronic transaction.
The core advantage was already coordinated participation: an institution could serve its own customers while using a network accepted beyond that institution.
Europe had already been acquired
Visa Europe completed its combination with Visa Inc. in June 2016.
That preceded Kelly's December CEO start. [S20][S75]
He inherited the combined organization and its operating obligations, rather than originating a completed acquisition from the CEO office.
The distinction matters commercially.
Buying a regional business and making a combined system work are different management jobs.
The growth constraint
More digital commerce created opportunity, but not every new transaction would look like a conventional card purchase.
Businesses needed cross-border transfers. Fintechs needed access to account information and payment infrastructure. Clients also wanted services surrounding a transaction.
A strategy confined to defending the existing card interface would address only part of those needs.
3. Business Thesis
Stay useful beneath changing payment experiences
Kelly publicly described three growth priorities in January 2021: consumer payments, new flows and value-added services. [S100]
The combination explains the commercial direction.
The existing network remained important. Additional capabilities could make Visa useful when the client's problem involved moving money or operating a payment business rather than issuing another card.
Connect capabilities rather than own every customer
The Tink announcement supplies a concrete example.
Kelly described combining Visa's network with open-banking capabilities to support European consumers and businesses. [S99]
The mechanism was not turning every connected account into a Visa-owned account.
It was giving client organizations more ways to build services using permitted data and payment connections.
The broader thesis was infrastructure relevance: remain useful across more of the work surrounding money movement.
4. Strategic Bets
Bet 1 — Address bank-to-bank business payments
B2B Connect launched globally in June 2019 during Kelly's CEO tenure. [S63]
The problem was high-value cross-border business payment coordination, not buying a consumer product at checkout.
The service connected participating banks through a non-card payment workflow.
That changed the type of payment Visa could help organize without requiring the payer to use an ordinary Visa card purchase.
The risk was adoption and integration.
A bank needs a working route, operational processes and reconciliation. A global launch does not make every destination bank or corporate customer an active participant.
The observable result was a launched service addressing a distinct institutional use case.
Bet 2 — Acquire cross-border operating infrastructure
Currencycloud completed in December 2021. [S13]
It supplied infrastructure through which banks and fintechs could offer foreign-exchange and cross-border services.
The purchased capability sat behind clients' own offerings.
That made the business different from a Visa-branded consumer wallet competing for every end user.
The mechanism was buying an operating platform and relationships instead of recreating every connection from the beginning.
The trade-off was integration responsibility: owning the company did not remove local requirements, partner dependencies or implementation work.
Bet 3 — Add consent-based account connectivity
Tink completed in March 2022. [S11]
Open banking addressed a different starting point: a customer allowing an authorized service to access financial-account information or initiate supported payments.
The acquisition announcement said Tink would retain its brand, management and Stockholm headquarters. [S99]
Existing teams and technology were part of what Visa sought to use, not obstacles automatically removed by the transaction.
The risk was making that capability useful within a larger organization while preserving the trust and permissions on which access depended.
The result was a completed operating-business acquisition, not ownership of all the banks or accounts it connected.
Bet 4 — Continue investment through a demand shock
The pandemic reduced travel and disrupted cross-border activity.
In July 2020, Kelly said Visa was continuing investment in strategically important areas, including e-commerce, contactless payments, new flows and services. [S74]
The choice was to preserve future capability while managing a weaker current period.
The tension was immediate: expenses and capital commitments had to be supported while transaction-related revenue was under pressure.
The public record establishes that investment stance. It does not provide a separate realized return for every project.
5. How Those Bets Changed the Business
Product
The portfolio extended beyond card-network participation into business-payment coordination, foreign-exchange infrastructure and open banking.
Those capabilities served different jobs rather than being new names for the same card.
Customer
The relevant buyer could be a bank's payments team, a fintech building a service or an institution managing cross-border activity.
The consumer using the eventual application was not necessarily Visa's direct customer.
Revenue logic
The wider portfolio created opportunities to charge for services and infrastructure.
Transaction value moving through a client system was still not the same thing as Visa revenue.
Distribution
Clients could incorporate Visa-owned capabilities into their own products.
That could use existing institutional relationships, while still requiring implementation and a commercial reason to purchase.
Moat
The proposed advantage combined network participation with capabilities useful around other payment activity.
Its strength depended on reliability and client value, not only on the number of acquired brands.
Industry position
Visa moved further into supplying other payment businesses.
That broadened its role but also placed more of its expansion within regulators' view of competition between existing and emerging payment methods.
6. Resource Allocation
Buying capabilities
The acquisition accounting recorded $1.863 billion of consideration for Tink.
Currencycloud's $893 million total included the value of Visa's previously held interest. It was not all new cash paid at closing. [S16]
These commitments show resources moving into operating capabilities.
They do not establish that every expected synergy had already occurred.
Retaining technical organizations
The Tink plan retained an existing brand, management and base.
That directed resources toward a functioning organization with its own skills and connections, rather than only a collection of code.
The management task included making those skills accessible to more clients without assuming that corporate ownership alone completed the integration.
Balancing resilience and growth
The core system also required attention.
A payment business cannot treat reliability as maintenance to be deferred whenever a new growth opportunity appears.
Visa Europe's 2018 disruption made that obligation concrete. A business with many dependent participants must invest in recovery and communication as well as new features.
7. Business Model: Before vs. After
| Dimension | At the December 2016 CEO start | By the February 2023 succession |
|---|---|---|
| Value created | Established network connectivity and payment services | Additional institutional money-movement and account-connectivity capabilities |
| Core product | Existing global network and digital portfolio | Same core plus more adjacent operating platforms |
| Buyer | Issuers, merchant-side institutions and partners | Those relationships plus buyers of newer service capabilities |
| User | Consumers, merchants and institutional teams | Additional client teams building or running non-card services |
| Revenue logic | Network-related activity and services | Wider potential service and infrastructure fees around the core |
| Software role | Already essential to processing and digital commerce | Additional APIs and acquired operating platforms |
| Distribution | Existing financial-institution and partner relationships | More client-embedded offerings through established and acquired relationships |
| Scale | Already international | Broader portfolio, with adoption varying by capability and market |
| Company boundary | Visa Europe already combined | Currencycloud and Tink added; Plaid not acquired |
The change was an extension of an existing infrastructure business.
It was not a switch from no software to software, or from no international operations to a first global network.
8. What Worked / What Didn't
Operating capabilities were added
B2B Connect launched, and Currencycloud and Tink completed.
Those are concrete outcomes: an available service and two acquired operating businesses.
Their long-term contribution depends on use, execution and economics after completion. A transaction announcement is not a stand-alone measure of business success.
The Plaid route failed to complete
Visa and Plaid terminated their proposed $5.3 billion merger in January 2021 after the DOJ challenged it.
The litigation ended without the acquisition. The DOJ's competitive allegations and Visa's contrary position should not be presented as a trial judgment. [S10][S100]
The commercial consequence is straightforward.
Visa did not gain Plaid through that deal, and time committed to the proposed route did not produce ownership.
Kelly's termination statement explicitly cited the time that protracted litigation could take. [S100]
International exposure remained
For the quarter ended June 2020, cross-border volume excluding intra-Europe transactions fell 47% in constant dollars.
Reported net revenue fell 17% to $4.837 billion. [S74]
A broad network could not remove the effect of closed borders and reduced travel.
Growth in another digital activity was not an automatic replacement for the economics of every lost cross-border transaction.
Reliability was tested
The June 2018 Visa Europe partial outage caused widespread disruption.
The Bank of England directed implementation of independent-review recommendations and required an external progress assessment.
The Bank expressly said its action did not imply a regulatory breach and was not enforcement action. [S55]
The incident required technical and governance work alongside the expansion into new services.
9. Business Legacy
Kelly's Visa period left a broader portfolio around an established network.
The important structural connection is between participation in card payments and supplying capabilities used elsewhere in the payment system.
The period also shows that expansion has limits.
A target can remain independent. A regional outage can affect confidence in a much larger network. A travel shock can change transaction economics despite digital adoption elsewhere.
Those constraints are part of the business history, not exceptions to remove from a favorable leadership story.
The services and acquisitions were delivered by Visa's teams, acquired organizations and institutional partners. Kelly's corporate leadership did not make him their sole inventor.
Career Background
Before Visa, Kelly held leadership roles at American Express and served as its president.
He also led the New York–New Jersey Super Bowl Host Committee. GM appointed him to its board in September 2024. [S28]
These roles provide operating context. The Visa analysis turns on payment-system decisions and outcomes rather than a catalog of outside affiliations.
Career Timeline
- 2016:
appointed CEO-designate in October; assumes full CEO responsibility in December.
- 2018:
Visa Europe experiences the partial authorization-system outage.
- 2019:
becomes board chair; B2B Connect launches globally.
- 2020:
pandemic disruption materially reduces cross-border activity.
- 2021:
Plaid agreement terminates; Currencycloud acquisition completes.
- 2022:
Tink acquisition completes.
- 2023:
McInerney becomes CEO; Kelly becomes executive chairman.
- 2024:
retires from Visa's board; later joins GM's board.
Common Misconceptions
“He acquired Visa Europe as CEO.”
The transaction completed before his full CEO start.
“Visa bought Plaid and later replaced it with Tink.”
The Plaid transaction did not complete. Tink was a separate acquisition with its own business and geography.
“New payment services eliminated dependence on travel.”
The pandemic results show that cross-border activity remained economically important.
“He remained CEO until his board retirement.”
The CEO succession and board retirement were separate events.
“Every product in the portfolio was invented during his tenure.”
Visa inherited technology and acquired teams with their own histories.
Evidence
- [S75], [S76], [S25], [S26], [S28] Role sequence and career.
- [S20] Visa Europe completion before the CEO appointment.
- [S63], [S13], [S11] Launched and acquired capabilities.
- [S99], [S100] Publicly stated strategy and the terminated route.
- [S16] Actual consideration conventions.
- [S74], [S55] Demand shock and operational failure.