Top 10 Lessons: What Launching a Neo Bank Taught Me About the Real Fintech Game
Part 02
Part 01 here:https://kanpurtocupertino.com/p/top-10-lessons-what-launching-a-neo
5. “Free” accounts are never free — to you.
Every account you publish as free still carries real backend cost: sponsor bank fees, BaaS charges, plastic card costs, ACH fees, faster-payment fees. None billed to the customer; all very real on your P&L.
This is exactly why the interchange-first mindset matters. Subsidize a free product with membership revenue or some adjacent stream that doesn’t scale with usage, and you’re in difficult terrain. The fintechs that thrive are the ones whose core engine grows in lockstep with the behavior they’re trying to encourage.
6. Incentives open accounts. They do not move balances.
We ran incentives, and they worked — at the top of the funnel. Sign-up bonuses absolutely get more people to open an account. The acquisition charts look great.
Here’s the trap: daily average balance barely moves until the product itself has reached a real MVP — something good enough to actually live in.
Incentives buy you an account opening. They don’t buy you a primary account.
As balance plus engaged usage are what feed interchange; mere incentive-driven signups without early interchange strategy is a cost, not a customer. Spend on incentives only once your product is ready to retain. Spend before that, and you’re paying full price for users who never actively engage.
7. Issue the plastic card from day one. “Digital-first” was penny wise, pound foolish.
We launched digital-first. The card lived in the app immediately; users could order plastic if they wanted it. The reasoning felt responsible — physical cards cost money, so let the people who need plastic opt in and save the expense on everyone else.
It was a mistake.
Our A/B tests said so plainly: the cohort that received a plastic card unprompted swiped more, more often, and stuck around longer than the opt-in cohort by a margin too large to attribute to noise. There’s something about the physical object — it sits in the wallet, it gets reached for, it keeps your brand top of mind in a way a card three taps deep in an app never will. Even with mobile wallets surging, debit users still reach for physical plastic for the majority of card-present purchases. That’s interchange you’re quietly forfeiting by making the card opt-in.
And with interchange as your North Star, the cost math inverts: not shipping plastic is the expensive choice. The card fee is a rounding error against an engaged user’s lifetime interchange. And if someone never activates the card you sent? That’s not a loss — it’s a re-engagement hook. You can nudge them, remind them, pull them back. You can’t re-engage a card you never put in their hands.
Send the plastic. Provision the digital card too, for instant use. But don’t make the physical card the thing users have to go find.
8. Wallet provisioning belongs in the MVP, not the roadmap.
If I had to name the single feature that punched above its weight, it’s mobile wallet provisioning — and Apple Wallet specifically.
The logic is simple: a card in a drawer generates nothing; a card living in someone’s phone gets tapped, and every tap is an interchange event. In the US, that phone overwhelmingly means Apple Wallet — Apple Pay dominates in-store mobile wallet usage by a wide margin. Provisioning into the wallet at activation is the difference between a card your user has and a card your user uses. Only the second one pays you. The friction of pushing a card to the wallet during onboarding is tiny; the engagement lift on the other side was the largest single onboarding change we shipped that quarter. Make it part of the MVP.
9. Don’t kick the can on strategically important decisions just because they aren’t urgent yet.
This is the one I’d most want to warn my earlier self about, because the failure mode is invisible until it isn’t.
There’s a category of decisions that feels safe to defer — not because it’s unimportant, but because nothing is forcing it today. Account closing and reopening strategy, ACH hold strategy, dynamic Payment journey limits, Bill pay integrations, ATM network additions. None of these is screaming at you during launch, so they quietly slide to “we’ll figure it out later.”
Then later arrives faster than you planned, and all at once. A user wants to close an account and you have no clean process. Your transfer hold strategy is suddenly the thing standing between a customer and their money. A partner asks about bill pay and you realize you never scoped it. Each of these, unaddressed, becomes a blocker the moment it’s relevant — and because you deferred the thinking, not just the building, you’re now making a consequential call under time pressure instead of with a plan.
The cost isn’t just the scramble. It’s the value you couldn’t deliver while you were untangling a decision you could have made calmly months earlier. Deferred strategy doesn’t disappear; it compounds into drag.
So separate “not urgent” from “not important.” The decisions above aren’t roadmap items you slot in when convenient — they’re foundational choices that shape what you can build on top of them. Make the calls early, even lightly, even knowing you’ll revise. A rough strategy you can refine beats a blank space you have to fill in a panic.
10. Customers open the account, then wait. Earn the switch.
People rarely switch their primary bank on day one. They open your account, fund it lightly, and quietly test you — does the card work everywhere, do payments clear, does support respond, does anything break?
Only after you pass that silent audit do they start routing their paycheck, their bills, their daily spend through you. That’s the moment you win, because that’s the moment interchange begins. So design for the test. Make the first few interactions flawless. Patience on the user’s side has to be met with relentless reliability on yours.
The part nobody puts in these posts
The honest footnote to all of this: the interchange-first model I’m preaching is also the model behind a long list of neobanks that died anyway. There is a graveyard of companies that nailed the thesis and still ran out of road — because the engagement never compounded fast enough to outrun the cost of “free,” or because acquisition got more expensive than a swipe could ever repay.
So I won’t pretend these lessons are a map to a guaranteed outcome. They’re the things that, in hindsight, separated the quarters where our unit economics moved in the right direction from the quarters where we were busy polishing the costume. Consumer fintech is an engagement business wearing a banking costume — and the costume is the part that’s easy to fall in love with.
If you’re building in this space right now, I hope a few of these scars save you some of your own.
What did your launch teach you? Especially if it contradicts something above — I’d rather hear that.