Skip to main content

Consumer payments

November 16, 2026

 

Getting to yes: Inside the art of turning attempts into approvals

How AI, network insights and payment optimization tools are injecting more intelligence into the approval process.

google logo

Vicki Hyman

Director,

Global Communications,

Mastercard

The payments industry has solved the speed question. It can take a fraction of a second to determine whether or not a transaction should be approved. But faster decisions don’t necessarily mean better decisions.

Every day, millions of legitimate payment attempts are declined — a tiny percentage of the billions of transactions processed each year, but each one represents a frustrated consumer and a lost sale — or worse: More than four in 10 consumers say they would never shop with a business again after a false decline, according to one study that also found that $50.7 billion was lost in 2022 due to false declines across the U.S., the U.K., France and Germany.

The problem isn’t the payment itself but the limited information available at the moment the transaction is evaluated — data that might be incomplete, outdated or not formatted in a way that meets processing best practices.

Does the customer have a card stored on the merchant’s site, or is this their first purchase? Has the card recently expired, with credentials that simply need to be updated? A recurring payment may come due at a time when available funds are temporarily low.

A failed payment can mean lost revenue and unnecessary friction for the merchant. For the consumer, it could ruin plans for an evening of takeout and streaming — or the loss of access to a critical business tool when on an even more critical deadline. For small businesses or solo entrepreneurs, that could be devastating.

 

We don’t want to assume that someone no longer wants our product if we don’t get a successful payment. We do everything we can to save a transaction.

Matt Wegner, Adobe

 

“We don’t want to assume that someone no longer wants our product if we don’t get a successful payment,” says Matt Wegner, head of Global Payments and Risk at Adobe. “They’ll have a lot of valuable content built in and saved, so likely they don’t want to lose that. We do everything we can to save a transaction.”

Now AI and network intelligence can bring the critical pieces of information together to create the context needed to make better decisions, increasing the likelihood that legitimate purchases will be approved.

"Every transaction tells a story,” says Tine Fincioen, executive vice president for Optimization at Mastercard. “The clearer and more complete that story is, the better the chances of reaching the right decision.”

While merchants only see their own data, Mastercard sees patterns across the network. Last year, Mastercard introduced its Payment Optimization Platform, which trains its AI model on its network of more than 175 billion annual transactions, continually analyzing and learning from patterns. It includes Payment Alerts, which helps subscription merchants decide the optimal time to retry a payment after a decline, and its newest development is Adaptive Authorization, which enriches the transaction information on behalf of merchants and acquirers in the split second before the data is sent to the customer’s card issuer for decisioning.

Using AI and machine learning, Adaptive Authorization analyzes signals from each transaction alongside broader authorization behavior observed across the network. It can take action to increase the likelihood the issuer will approve the transaction — updating expired card credentials or improving how transaction data elements are populated to support proper processing, for example — all in real time.

The result: More legitimate payments get approved, helping merchants close sales and eliminating friction for the cardholder.

 

From smarter approvals to smarter retries

Payment Alerts applies its intelligence after a recurring payment has been declined for financial reasons — when a debit account has insufficient funds or a credit card has reached its limit.

For Adobe, the challenge isn’t simply whether to try the payment again. It’s knowing when. Retry too soon and the payment may fail again. Retry too often and merchants might face additional costs or signal to issuing banks they’re not managing their risk correctly, which could contribute to even more declines.

“We may take 30 days or longer to save a transaction,” Wegner says. “We want to be really thoughtful about saving an account, but not at all costs. We want to be smart about how we do it.” 

 

Introducing Performance Pulse

Getting a transaction approved is key to making payments work better, but strong card performance extends beyond checkout. For banks, the opportunity spans the customer relationship, from keeping customers engaged to strengthening long-term portfolio performance.

 

Mastercard research shows retaining an existing customer can be up to seven times more cost-effective than acquiring a new one, while reactivation programs can generate approximately 5% uplift and stronger retention can improve profitability by as much as 75%.

 

In Q1 2027, Mastercard will launch Performance Pulse to help banks understand portfolio performance and prioritize where to act. Built into the Mastercard Intelligence Center, and grounded in Mastercard’s transaction data as well as bank-reported data, Performance Pulse will bring together diagnostics, benchmarking, recommendations and pathways to action to help banks identify opportunities for greater engagement and growth.

 

Performance Pulse will launch first in the U.S., with an initial focus on driving spend activation for existing customers, followed by a global rollout later in 2027, with a broader suite of portfolio performance capabilities.

 

That’s what made a different kind of signal particularly intriguing: What if the company could know that a card that had failed at Adobe had subsequently worked somewhere else?

With its Payment Alerts capability, Mastercard can determine the best time to retry for a successful payment authorization.

Here’s how it works: In near real time, the Payment Optimization Platform monitors the Mastercard network for financial declines to pass on to its model for further analysis. The model can identify patterns on the credential to determine the optimal chance of approval — for example, when it has been used successfully several times — and once the credential’s real-time activity matches the model’s recommendation, an alert is sent to the merchant to retry the transaction immediately.

“It has become one of the strongest retry strategies in our entire portfolio of tools,” Wegner says.

This approach grew out of conversations between Adobe and Mastercard about whether network-level intelligence could solve a problem Adobe struggled with when relying solely on its own data. Wegner had shopped this idea around to several players in the payments ecosystem and got some hard no’s, but Mastercard opened its product development process to him. Once teams from both companies found a way to provide the signal Adobe needed without exposing sensitive information, Adobe piloted the capability and offered feedback to Mastercard on how it could be expanded to other subscription merchants.

Together, these capabilities address different points of friction, but they’re all designed to produce the same result — so a customer’s legitimate purchase doesn’t fall through the cracks.

“The best payment experience is one you barely notice,” Fincioen says. “It just works. The more often we can make that happen, the better it is for everyone.” 

As a new payments paradigm evolves, we are grounding it in trust

As AI agents and tokenized currencies reshape commerce, Mastercard is embedding security, governance and user protections so new payment models can scale responsibly.
A woman touches a glowing screen with one finger.