Your browser does not support JavaScript! Please enable the settings.
↑

After Launch: The Operating Model for a Card or Wallet Programme

Most brands plan and fund the launch of a card or wallet and then discover the run. This guide sets out the day-two operating model: tracking scheme changes, handling disputes, reconciling money, supporting customers and resourcing the whole thing.
November 16, 2026
Innovify Editorial
published on
November 16, 2026
After Launch: The Operating Model for a Card or Wallet Programme

After Launch: The Operating Model for a Card or Wallet Programme

Brands tend to budget carefully for the launch of a card or wallet and far less carefully for what follows. The launch has a date, a project team and a definition of done, while the programme that follows has none of those. Scheme rules change on schedules you do not set, customers dispute transactions on days you did not plan for, and the money in your ledger has to agree with the money held elsewhere every day. This guide describes the day-two operating model, from scheme changes, disputes and reconciliation to support and resourcing, for a Head of Embedded Finance, Chief Digital Officer or CTO launching a financial product inside a brand that is not itself a bank.

Launch is a milestone, not the programme

A card or wallet programme is a chain of parties, each carrying part of the work. Most programmes involve some combination of the brand that owns the customer relationship, a regulated partner that holds the permissions, an issuing or wallet platform, a processor, the card schemes and one or more banks. Which party is responsible for which duty depends on your contracts and regulatory structure, so confirm it with your regulated partner and your own advisers.

What does not change under any structure is that the customer sees one brand. When a card is declined, a refund is late or a balance looks wrong, the customer contacts you. That is the first principle of the operating model: you own the customer experience of the programme even where you do not own the underlying activity. The work of running the programme is therefore partly doing things and partly managing the parties who do them.

It helps to think of five operating loops, each with an owner, an input, an output and a cadence:

  1. Change management. Tracking and absorbing changes from schemes, processors and regulators.
  2. Dispute handling. Taking a customer's complaint about a transaction through to an outcome.
  3. Reconciliation. Proving that records, partner files and held funds agree.
  4. Customer support and incident response. Answering people, and handling things that go wrong at scale.
  5. Oversight. Reporting, controls testing and review with your regulated partner.

If you are scoping a programme with Innovify or anyone else, ask early how these five loops will be staffed. Our embedded finance and digital wallets work starts from the view that the operating model is part of the product, not an afterthought to it.

Scheme changes: the work nobody scheduled

Card schemes publish rule and technical changes on a recurring cycle. Processors and issuing platforms then pass those changes to programme owners, usually as notices with effective dates. Some are informational. Some are mandatory and require action. The danger is that a change which looks minor on paper touches something your product depends on.

One example from the partner evidence we hold illustrates the shape. A card-issuing platform's published change notice, announced on 29 September 2026, listed Visa network changes taking effect on 23 October 2026. According to that notice, as summarised in our partner monitoring, they covered an expanded authentication indicator for card-not-present transactions in further regions, and new merchant category codes appearing in transaction payloads, webhooks and records. We have this from a single partner source and have not checked it against the scheme's own documentation, so treat it as an illustration of the kind of change that arrives, not as guidance on that change. We have deliberately not named the platform.

Our reading of why such a change matters, which is analysis and not reported fact, runs as follows. A new category code can break a rule that blocks or allows spending by merchant type. A new field can break a parser that expects a fixed payload. A changed authentication indicator can alter how your fraud rules or reporting interpret a transaction. None of that is dramatic, but all of it can silently degrade a product that worked on launch day.

A change register that works

  • Intake. Name the people who receive notices from each partner and scheme source, and a shared mailbox or channel where they land. A notice that arrives in one person's inbox is a notice that gets lost when they go on leave.
  • Triage. Within a set period, classify each change as informational, mandatory or product-affecting, and record the effective date.
  • Impact assessment. For anything mandatory or product-affecting, ask which rules, parsers, reports, limits and customer messages depend on the thing that is changing.
  • Test. Use the partner's sandbox or a staging environment before the effective date. If a change cannot be tested before it takes effect, record that as a risk.
  • Release and evidence. Record what was done, by whom and when, so you can show your regulated partner a closed loop.

Someone should own the register and review it regularly with your processor or platform contact. Where a notice is ambiguous, ask the partner in writing.

Disputes: a flow with deadlines you do not control

When a customer disputes a card transaction, a clock starts. The detail of timelines, evidence requirements and outcomes is set by scheme rules and your agreements, and varies by scheme, transaction type and region. We do not state specific periods here. Confirm them from your processor and scheme documentation.

What you can design is the flow. A workable dispute process has these stages:

  1. Intake. Capture the customer's account of what happened, the transaction identifier and the reason, in a consistent form. Poor intake is the most common reason evidence is late or incomplete.
  2. Classification. Decide whether this is a fraud claim, a merchant dispute, a processing error or something that is not a dispute at all, such as a customer who does not recognise a merchant's trading name.
  3. Evidence gathering. Collect what the scheme process requires, from your own systems, your processor and, where relevant, the merchant.
  4. Submission and tracking. Submit within the deadline and track the case to a decision.
  5. Outcome and ledger treatment. Apply the result to the customer's balance and to your own records, and tell the customer what happened.
  6. Learning. Feed patterns back to fraud controls, product design and customer communications.

Two design points matter. Decide in advance who may approve a goodwill credit and up to what limit, so support staff are not improvising. And link disputes to ledger movements in your records, so reconciliation can explain every adjustment.

Wallets deserve a separate note. A stored-value wallet, a card linked to a wallet and a card on its own can each carry different dispute routes and different customer rights. Which applies depends on your product design and the permissions behind it. Ask your regulated partner and legal advisers to set that out in writing before launch, as this is not legal advice.

Reconciliation: proving the money agrees

Reconciliation is the operating loop most likely to be underestimated, because when it works nobody sees it. In outline, you are comparing three views of the same reality: what your own ledger says each customer is owed or owes, what your processor and scheme files say happened, and what the bank accounts holding the money actually contain.

Differences are normal. The skill is classifying and clearing them promptly. Usual categories are timing differences between authorisation and settlement, fees and interchange, refunds and reversals, currency conversion, chargebacks and their outcomes, and genuine errors.

A sound operating model for reconciliation includes:

  • A daily routine with a named owner and a named deputy. Weekly is too late to learn that a file failed on Monday.
  • An exceptions queue with ageing. Every unmatched item has an owner, a reason code and a date. Items that age past a threshold are escalated.
  • Break categories that map to causes. If most breaks are timing, you have a different problem from a programme where most are unexplained.
  • Evidence retained for review. Your regulated partner and auditors will ask how you know the money agrees.

For wallets and other stored-value products, the way customer funds are held and protected is a regulatory matter that depends on your structure and your partner's permissions. We describe it here only in outline. Read the FCA's current guidance for your product type and take advice before you design the money-movement and reconciliation flows.

Support and incident response

Support for a financial product is not the same as support for a retail product. The customer's money is involved, so contacts carry urgency and sometimes distress. Plan for these from the first week:

  • Card controls. Freezing, unfreezing, reporting lost or stolen cards and replacement, available at any hour a customer might need them.
  • Declined and disputed transactions. Staff need to see enough of the transaction record to explain what happened, and a route to escalate to the processor.
  • Fraud and scam contacts. A defined path to the right team, with clear rules for what front-line staff may and may not promise.
  • Complaints. A recognised complaints process, with the timelines and recording that apply to your structure, agreed with your regulated partner.
  • Customers who need extra help. Our reading is that the Consumer Duty, which the FCA applies to firms in scope, makes outcomes for customers, including those in vulnerable circumstances, a live operating concern. Whether and how it applies to you depends on your structure, so check with your regulated partner.

Incident response needs equal attention. Decide beforehand who declares an incident, who speaks to customers and partners, and how you will say what is happening while the cause is still unknown. A processor outage on a busy trading day is not the moment to draft the template.

Resourcing the run

Brands often ask what a programme costs to run. We do not quote figures, because they depend on volumes, structure, partners and scope. What can be said is where the cost sits, so you can ask partners the right questions:

  • People. Programme management, operations, reconciliation, disputes, support, finance and risk. In a small programme these are part roles, but each still needs a named owner.
  • Partner charges. Per-transaction, per-card, per-account and service fees, plus charges for disputes, replacements and ancillary services. Ask for the full schedule, including events that happen only occasionally.
  • Technology. Integration upkeep, monitoring, reconciliation and case-management tooling, and engineering time to absorb scheme and platform changes.
  • Compliance and oversight. Reporting, controls testing, audit support and time with your regulated partner.

Our reading is that a common budgeting error is to treat the integration as finished at launch. It is live infrastructure. Platforms and schemes change, and each change costs attention if not always money. A useful discipline is to fund a standing capacity for change, adjusted by what you learn in the first quarters.

On the question of who does the work, there are three broad options. You can run it in-house, which gives control and requires hiring. You can use managed operations from a partner, which shortens the ramp-up and requires careful service levels and oversight. Or you can blend the two, with your team owning customer and product decisions while a partner handles defined operational loops. None of these is wrong in general. The mistake is to choose by default, after launch, under pressure.

Where Innovify fits

Innovify is a technology partner with deep expertise in banking, financial services, fintech and insurance, and an implementation partner across multiple fintech platforms. We are not a regulated financial institution, and regulated activity stays with you and your regulated partner. What we can do is help a brand treat the operating model as part of the build: designing the change register, the dispute and reconciliation flows, the support tooling and the monitoring alongside the product, with compliance designed in from day one.

That is Innovify's stated approach. If you are planning a programme or already live and finding that the run is heavier than the launch, our Embedded Finance & Digital Wallets team can help you map the five loops against your own structure.

FAQ

What does it take to run a card or wallet programme after launch?

It takes five operating loops with named owners: change management, dispute handling, reconciliation, customer support with incident response, and oversight. Each needs a routine, a person accountable and evidence that it ran. Which parts you do yourself and which a partner does depends on your contracts and regulatory structure.

Who is responsible for scheme rule changes?

Scheme rules are set by the schemes, and processors or platforms usually relay changes to programme owners, but responsibility for acting on them depends on your agreements. Treat a notice as your responsibility to triage unless a contract clearly says otherwise. Keep a change register so that every notice has an owner, an impact assessment and a record of what was done.

How should disputes and chargebacks be handled?

Handle them as a defined flow: intake, classification, evidence gathering, timely submission, outcome applied to the ledger and learning fed back to controls. Deadlines and evidence rules come from the schemes and your processor, so confirm them from those sources. Decide in advance who can approve goodwill credits and up to what limit.

Why does reconciliation matter so much for a card or wallet?

Because it is how you prove that your ledger, your partners' records and the money actually held agree. Differences are normal, often timing, fees and refunds, and the discipline is to classify and clear them daily. Unexplained breaks that age without an owner are an early warning of larger problems.

What does it cost to run a card programme after launch?

It depends on volumes, structure, partners and scope, so no single figure is reliable. Costs sit in people, partner charges, technology upkeep and compliance oversight. Ask each partner for a full fee schedule, including occasional events such as disputes and replacements, and fund standing capacity for scheme and platform changes.

Conclusion

A card or wallet programme is a service you operate, not a project you finish. The launch proves the product can work. The operating model decides whether it keeps working as schemes change, customers dispute, money moves and incidents happen. Name the owner of each of the five loops, build a change register before you need one, and design the dispute and reconciliation flows alongside the product. Verify scheme details with your processor and the scheme documents, and regulatory points with your regulated partner and advisers.

Related reading

Explore next: Answers to common questions

Connect With Us

Let’s Build Something Exceptional

Innovify transforms ideas into scalable, impactful products by integrating user centered design, AI-native engineering, data-back growth and lean agile methodologies. Our close client collaboration complemented with on-shore and nearshore delivery creates speed, clarity, and success at scale.
Discuss your project