Netflix knows what you want to watch before you do. Spotify builds playlists around your mood. Amazon predicts what you need before you run out. We live in an era of hyper-personalisation, where the platforms we choose to spend money with have become unnervingly good at knowing us.

Now consider the companies we have no choice but to pay. The utility provider. The insurer. The healthcare system. The tax authority. These are the payments that carry the most emotional weight, the ones tied to keeping the lights on and the family covered, and yet they are the ones most likely to treat us like an account number.

“Customers are tired of being treated like strangers by their service providers to whom they send money every month,” Nicole Haskins, Director of Customer Experience at Paymentus, told PYMNTS.

That gap between how we are treated by the services we want and the services we need is not a minor UX issue. It is a strategic failure. And fixing it may require the payments industry to accept an uncomfortable truth: the last decade of innovation solved the wrong problem.

The first chapter of modern payments was about access. Moving from paper to pixels. Cheques to cards. Manual invoicing to electronic billing. Cash registers to point-of-sale terminals.

The metric that mattered was coverage. Could a merchant accept a card? Could a biller send an electronic statement? Could a consumer pay online without posting a cheque?

This chapter took decades and it reshaped the global economy. Visa and Mastercard built networks spanning more than 200 countries and territories. PayPal made online payments possible for a generation of early ecommerce adopters. Mobile wallets extended the reach to markets where bank accounts were scarce but smartphones were not.

The digitise chapter is largely complete. The infrastructure exists. In most developed markets, the question of whether a payment can happen digitally has been answered. The rails are laid.

But digitisation was never the destination. It was the foundation.

The second chapter is the one the industry is still writing, and it has produced extraordinary gains.

Faster rails. Cheaper processing. Smarter fraud detection. Stripe processed $1.9 trillion in payment volume last year. Real-time payment networks like FedNow and Faster Payments are shrinking settlement times from days to seconds. Predictive analytics flag suspicious transactions before they clear. Checkout has become invisible: one tap, one click, no friction.

Artificial intelligence is accelerating the optimisation race. Visa leads the payments industry’s AI adoption, with Mastercard and PayPal close behind, according to a recent global benchmark. The three are investing heavily in AI-driven fraud prevention, personalised offers, and automated dispute resolution.

Mastercard’s Mark Barnett, Global Head of Small and Medium Enterprises, reinforced this in a PYMNTS interview this week. His message to small businesses was blunt: growth comes from using existing technology better, not chasing trends. Integration, not innovation, is the bottleneck. Administrative fragmentation represents “a structural tax on time” for business owners whose reservation software does not talk to their inventory system and whose payments do not auto-reconcile with their accounting.

The metrics of Chapter Two are uptime, throughput, cost per transaction, and conversion rate. By those measures, the industry has delivered. Payments are faster, cheaper, and more reliable than at any point in history.

But optimisation has a blind spot. In the race to remove friction, the industry also removed something it did not intend to: the sense that there is a human on the other side.

This is the chapter that Haskins is arguing the industry must now write.

Her word of the year is “service.” Not a new protocol. Not a faster rail. Not another AI model. Service.

The argument is straightforward. Billing interactions are fundamentally different from discretionary purchases. You choose to subscribe to Netflix. You do not choose to pay your electricity bill. One is entertainment; the other is survival. When money gets tight, the emotional stakes of these mandatory payments intensify.

“As money and resources get tight, service can make or break a relationship,” Haskins told PYMNTS.

Yet the personalisation gap between discretionary and essential services is stark. Streaming platforms know your preferences down to the genre and the time of day. Your insurer sends you a generic PDF. Ecommerce sites remember your size, your style, and your shipping address. Your utility provider cannot tell whether you are a homeowner or a tenant, whether you have been a customer for two months or 20 years.

Traditional fintech metrics do not capture what is being lost. Uptime does not measure whether a customer felt heard when they called about a billing error. Throughput does not reflect whether a payment reminder arrived at the right moment or the worst possible one. Cost per transaction says nothing about whether the experience built loyalty or eroded it.

The grocery sector offers a parallel. PYMNTS reports that the battle between Walmart and Amazon has shifted decisively to loyalty. Walmart captures 72 percent household grocery penetration. Amazon grew perishable delivery sales 40 times since January 2025. Both are competing on experience and habit formation, not just price. Grocery is the closest thing retail has to a recurring bill, and the companies winning that fight are the ones that make repeat interactions feel effortless and personal.

The same logic applies to every biller, insurer, and service provider that collects a monthly payment. Win the experience and you earn the relationship. Fumble it and the customer leaves the moment a competitor offers a smoother path.

The answer is not to undo the gains of Chapters One and Two. Nobody wants to go back to posting cheques or waiting three days for a payment to clear.

The answer, Haskins argues, is “flexibility and optionality.” Let users choose. Some customers want pure self-service: a clean portal, a clear bill, a one-tap payment. Others want to speak to a human when something goes wrong. The mistake is forcing everyone down the same funnel, usually a chatbot, and calling it innovation.

This tension is already playing out in one of the industry’s fastest-moving frontiers. As we explored in our analysis of the delegation problem in agentic commerce, AI agents are beginning to transact on our behalf. Stripe, Visa, and a growing ecosystem of startups are building the infrastructure for machine-to-machine payments. The question of when an automated system should hand back to a human is no longer theoretical. It is a design decision with commercial consequences.

Chapter Three does not reject automation. It demands that automation be designed with the awareness that a person exists at the end of the chain. The bill is not a data packet. It is a moment of trust.

The payments industry spent a decade digitising and another optimising. Both chapters created enormous value. But they also created a gap: a generation of payment experiences that are technically flawless and emotionally hollow.

Chapter Three asks a different question. Not “how fast can we process this?” but “how does this interaction make the customer feel?” Not “can we automate this touchpoint?” but “should we?”

The billers, processors, and platforms that answer those questions honestly will earn something that uptime and throughput never could: loyalty in categories where switching costs are low and frustration is high.

The companies that treated payments as solved plumbing will discover that the pipe was never the product. The relationship was.

Will the payments industry remember that behind every recurring transaction is a person who wants to feel known?