Hidden Financial Services: What Business Owners Should Learn

Financial services are increasingly being designed to work in the background. Customers may use a familiar app to move money or shop with an AI agent without managing the systems behind those actions.

That was a central theme at Money20/20 USA, where more than 11,000 payments, fintech and business leaders gathered in Las Vegas. In a Stripe account of the event, speakers described three developments: stablecoin wallets embedded in existing products, financial services that operate around the clock, and clearer controls for payments made by AI agents.

These developments are not a checklist every business should rush to adopt. They are a useful prompt to ask how money, customer experience and automation might fit together in your business—and what customers need to understand and control.

Financial services can sit inside familiar products

One approach to stablecoins is to place wallet functionality inside an app or website people already use. That can look different from asking customers to download a separate crypto wallet. The product might let a user receive a stablecoin-powered payout without presenting the experience as a crypto service.

Amira Valliani of the Solana Foundation described two models: dedicated crypto wallets and “hidden” or white-labeled wallets built into another product. A mobile-device manufacturer, for example, could add international money transfers to the wallet app already available on its phones. The customer would interact with the familiar app, while the financial capability sits behind it.

For a business owner, the broader lesson is about reducing unnecessary friction. If a financial feature supports a real customer need, think about whether it belongs in a journey customers already understand. Avoid adding a complicated new destination when an existing product or service can present the feature more simply.

That does not mean hiding important information. A smooth experience should still make it clear what a customer is doing, what the service costs, and how funds move. The source describes the value of simplifying the infrastructure, not removing customer understanding or choice.

Before exploring an embedded financial feature, write down the customer problem it is meant to solve. Then map the steps from discovery to completion. Identify where customers might get confused, and decide what information or support they need at each step.

Always-on money can change when services are available

The second theme was money movement outside traditional banking hours. The source describes stablecoin settlement as enabling financial services to respond to demand in real time, including outside those hours.

Citi Treasury and Trade Solutions gave an example involving a broker-dealer seeking to buy US equities during Chinese New Year, when banks in Singapore were closed. With tokenized deposits, the company could move funds when needed instead of forecasting its requirements and sending money to the US in advance.

Other examples showed different uses. Valinor Digital used stablecoins for a financing facility that lets prediction-market businesses access capital in real time, including overnight and on weekends. Robinhood used on-chain lending markets to let customers earn potentially higher returns on stablecoins than through its traditional cash product. Airtm uses stablecoins to help digital workers outside the US and Europe receive earnings from companies in those markets, then connect those funds to local options for saving, spending and withdrawing.

These examples do not mean every business needs stablecoins. They do show why owners should examine their own operating hours and cash-flow processes. Do customers need to pay, receive funds or access a service when your team or usual financial providers are unavailable? Do you spend time forecasting and prefunding accounts? Those questions can help identify where a faster or more flexible process might matter.

Start with the workflow, not the technology. Document when a payment or funding need arises, what currently happens, and where delays or manual steps occur. Then compare possible approaches against the actual need. Consider who manages the process, what the customer sees, and how exceptions are handled before deciding whether a new financial rail is relevant.

AI payments need clear limits

The third development concerns AI agents that can make purchases on a customer’s behalf. A payment credential can show an agent has access to funds. By itself, though, it does not explain what the customer authorised the agent to buy.

Natural-language requests can be open to interpretation. Someone might ask for cheap headphones that arrive quickly. The agent still has to interpret “cheap,” “quickly” and “headphones.” Even a reasonable guess could result in a purchase the customer did not intend.

The article describes a more structured approach. An agent submits a spending request with details such as the amount, seller and purpose. The buyer can approve that request before the purchase is completed. The request information can also inform fraud models. Stripe’s Link wallet for agents offers purchase protections for eligible purchases, with specified protections for qualifying transactions.

For business owners, the takeaway is not simply to add an agent to checkout. It is to define what the agent may do and how a customer can review or approve its actions. Consider setting clear rules for eligible products, spending limits, sellers and permission expiry. The source specifically points to structured records of what an agent can buy, how much it can spend, and when its permission expires.

If you are exploring AI agents for authorised tasks, make the permissions part of the design from the beginning. Decide what actions require customer approval. Make it possible to see what the agent intends to do before it acts, rather than relying only on a broad instruction in natural language.

Make the customer experience the starting point

Across all three trends, the common idea is that customers should be able to use a helpful service without managing unnecessary infrastructure. But a business still has to choose which complexity to remove and which information customers need to retain.

Use a customer journey to test the idea. Follow the experience from the customer’s first question through to the completed task. Mark where they need to understand a new product, repeat information, wait for a process, or make a decision without enough context. Those points can reveal whether an embedded feature or automation is useful.

Then check the operational side. A new payment or AI-enabled process may connect several tools and teams. Before changing it, map the systems involved, the data they need to exchange, and the person responsible when something does not go as planned. Automation and integrations can help businesses think through how existing tools connect, while a website audit can help surface issues in the online experience where customers encounter your services.

Keep the first version focused. Choose one customer need or internal workflow to examine, rather than trying to redesign every payment process at once. Define how the customer should experience it and how your team will handle questions, errors and approvals.

What to do next

1. Find a real point of friction. Ask customers and staff where payments, access to funds or purchase steps feel confusing or slow.
2. Map the current process. Record each step, the systems involved and what happens outside your usual operating hours.
3. Check whether the trend fits. Embedded wallets, always-on money movement and agentic payments solve different problems. Do not adopt a technology simply because it is prominent.
4. Set customer controls. For any AI agent that can spend, define its scope, limits, approval steps and permission period.
5. Review the full experience. Make sure customers can understand the action they are taking and know where to get help.
6. Test a narrow change. Start with one well-defined customer or business need. Review what works and what needs adjustment before expanding it.

The Money20/20 discussion points to a practical direction: useful financial services may become less visible, but good customer decisions still depend on clarity. Start with the job your customer needs done, then decide whether a new financial feature or automated process genuinely helps.