Image by Simon from Pixabay

What makes save now, buy later (SNBL) a winner

Over the past half-decade, embedded finance and its supporting technology have matured to make save now, buy later (SNBL) a winning option, Ahon Sarkar said.

Sarkar is the general manager of Q2’s Helix division. Helix works with companies to make personalized banking products.

Sarkar said embedded finance had undergone several iterations in its short history. In 2017, many non-banking companies questioned the value of BaaS products, asking why they should offer banking when they’re not a bank. It then caught on, overheated, and seems to have settled in as a long-term, sustainable influence in the financial system.

Prioritizing people through SNBL

A key aspect of embedded finance, perhaps a responsibility, is to meet the needs of the underserved, Sarkar said. The world’s full of knockoffs and repeats. Technology exists to create entirely new things.

“What we’ve learned over time is that the magic of embedded finance is not having a bank account or debit card,” Sarkar said. “It is solving a problem in a particular context that makes that whole business better.”

SNBL is an example of building financial products around humans instead of shoehorning people around an existing product suite. Sarkar estimates that up to 20% of people aren’t well-served by financial products. It’s not their fault, and it’s not necessarily the banks’ fault, either.

Blame technology, he responds. It’s expensive, making low-deposit customers unprofitable unless they get charged high fees to cover the tech outlay.

Three embedded finance changes that paved the way

Sarkar said three key factors had to change for this new generation of embedded finance.

First is the business model. White labeling and pre-paid cards aren’t new. The problem is that one company’s brand is visible, but another institution runs it. That makes traditional core systems pricey propositions. Serving the neediest becomes cost-prohibitive.

Then capabilities had to change. In the past, banking technology was designed for banks to serve their customers. Designing products for banks to serve customers who serve millions more is more complex. Developments like multi-tenant cloud cores help.

Entrepreneurs need available whitespace to tackle the need, Sarkar added. They first must recognize there’s a problem and then have to work on solving it differently. In 2017 visionaries would approach companies that would most benefit from new solutions and were often told to prove the concept.

In doing that, Sarkar learned to look for companies that would see the landscape as he did. Those who did were one derivative away from banking, who understood the tech.

“Nobody would think about solving the problem using tools they didn’t know existed,” Sarkar reasoned. “It’s like Ford’s adage; if he asked his customers what they wanted, they said a faster horse. Nobody knew the car existed.”

Lessons from fintech’s burst bubble

As embedded finance boomed, the whitespace shrunk, but the taps kept pouring money. Investors were agog and saw revenue and user growth. Investors saw a market opportunity that fuelled a surge in companies and products. In a frenzy, some companies just launch anything, and it becomes a race for progress so they can raise another round.

Reality began to set in in early 2022, Sarkar said. Upon closer inspection, customer growth was sometimes driven by free products but was costly to the company. Founders saw their business plan was well-executed but was based on losing more and more money over time. Many were faced with gutting their product.

Sarkar’s take? Focus on quality, not quantity. His goal is to serve the entirety of the United States through the fewest number of customers. Make unique products for unique people.

Start with solving a specific problem. Then solve another one. In the case of SNBL, that means helping companies generate profitable growth while maintaining their core business.

How those lessons apply in SNBL

Appropriately designed, SNBL solves problems for businesses and their consumers simultaneously, Sarkar explained. Why were companies willing to pay the interest on BNPL loans? To curb customer abandonment, he said. Give them the money, pay the interest, and hope for the best.

Over the past year, the water’s flowed out to sea, and more than a few weren’t wearing their trunks. Count both providers and customers among them. The most vulnerable consumers were susceptible to debt stacking.

Ahon Sarkar headshot
Begin with a human-centered approach allows fintechs to design effective SNBL products, Ahon Sarkar said.

Sarkar said that some doubled down and tried to cut costs, which hurt the whole ecosystem. Folks soured on the concept as interest rates rose.

But that brought opportunities. Now, there was money to be made in deposits. Combine that with solving a customer problem, and you have something.

The problem hasn’t changed, Sarkar observed. People still need the money to buy something they want. Perhaps it’s a mortgage. Maybe a trip or only a barbecue.

It’s also the same for the company. Both their customer acquisition costs and customer drop-off rates are too high. Many customers end up buying from someone else.

Can you solve the problem for the bank while offering the customer a faster path to their desired purchase? Sarkar can, in a way, makes the financial system a little more human.

Making it work

With embedded finance, how can you help companies pay more out than they receive if they pay out the interest in cash while not risking revenue? Consider a mortgage company wanting to help folks save for a down payment while giving them the value they can’t get elsewhere.

The solution? Give rewards equal to a certain cash percentage but only convertible to cash upon the transaction. If they stay, you eat the 1% because you make much more on the mortgage. If they leave, you’ve earned interest on the amount they saved up with you.

“You’re staying the same, getting free revenue, or lowering costs,” Sarkar said.

The consumer gets a rate they won’t get anywhere else. They also reach their transaction goal faster because the eventual transaction subsidizes it.

Too many BNPL companies scroll past 90 or 95% of the customers at the top of the funnel to focus on the five or 10% that make it to checkout. What if a company can design a simple and convenient product and move the conversion up five or 10 points?

This can be done without incentivizing bad behavior.

“Part of making finance human is trying to make people healthier,” Sarkar said. “This is one area where incentivizing people to save in a world where half don’t have savings, trying to make that easier while getting more out of their money. That is a worthwhile problem.”

When meeting with prospective clients, Sarkar’s team discusses the problem and how a solution could fit into their ecosystem. Helix’s complete API-based platform is also a core banking system, so there’s no intermediary. The Helix team adds the user interface inside of an existing app. Clients are matched to partner banks with deposit networks often spread across the country.

Sarkar said real-time payments would be a growing disruptive force. They will change how products are built and managed, leading to even more real-time systems and the decline of batching programs.

He reasoned that business accounts and lending are the sectors most ripe for disruption over the next two to five years. Most of their systems are built on legacy technology. Lending can be more than a melange of systems that fail to capture all the nuances of a customer’s financial life.

Then there’s the safe use of AI and the ethical questions it raises, Sarkar said. For example, companies need to categorize workflows to help new employees quickly get up to speed. Usually, they’d assign experienced staff to catalog each process, chewing up weeks of development time and thousands of dollars.

What if you could get ChatGPT within Azure Open AI to do it? Sarkar did, and instead of weeks and thousands of dollars, it took minutes and less than $10.

Documentation and quality assurance are easy targets. ChatGPT can now ingest images. What does that say for manual review?

But what’s the honest thing to do? Sarkar asked. On the good, it can take a meeting transcript, analyze it and create follow-up tasks. Soon you can tell it to build an entire game for you. Add data for a quality assurance script to evaluate it for you.

“Is it perfect? Absolutely not,” Sarkar said. “Think about the rate of change, which will impact the nature of work. I can imagine a world (where soon) the best engineers are the ones who know how to write the best prompts because writing a prompt is hard.

“What does that do to people who specialize in computer science? What does that do to how we as a society value work? Regulators think about this kind of thing’s impact inside the broader economy.”

Also read:

  • Tony Zerucha

    Tony is a long-time contributor in the fintech and alt-fi spaces. A two-time LendIt Journalist of the Year nominee and winner in 2018, Tony has written more than 2,000 original articles on the blockchain, peer-to-peer lending, crowdfunding, and emerging technologies over the past seven years. He has hosted panels at LendIt, the CfPA Summit, and DECENT's Unchained, a blockchain exposition in Hong Kong. Email Tony here.