I still remember the exact moment I realized order could be toxic.
It was 2019, deep inside the risk committee of a major regional bank. A tiny pilot project – letting customers round up debit-card purchases into a crypto savings bucket – had triggered a 47-page risk memo and three board-level escalations. The idea died quietly that afternoon. Everyone in the room felt relieved. I felt sick. Because in that moment I understood: we had become so good at preventing failure that we had accidentally made success impossible too.
That experience never left me. And it’s the reason I get oddly excited when people tell me “banks can’t innovate.” They absolutely can. They just forgot how.
The Hidden Cost of Being “Too Safe”
Banks aren’t lazy. They’re exhausted from carrying the weight of perfection.
Every process is audited. Every edge case is stress-tested. Every new idea has to survive a gauntlet of compliance, legal, risk, IT security, brand, and reputation teams before it even reaches a prototype. The system works beautifully – right up until the world changes faster than the system can approve change.
And the world is changing fast.
Tokenization of real-world assets is moving from experiment to trillion-dollar reality. Embedded finance is turning every software company into a potential banking competitor. AI agents are writing smart contracts that used to require entire legal departments. Yet most banks are still debating whether they’re allowed to run a sandbox with $50,000 of play money.
The more control you insist on, the less room you leave for discovery.
I’ve watched brilliant product managers quit because their ideas that would take three weeks in a fintech took three years inside a bank – if they survived at all. The talent doesn’t leave because the paycheck is bad. They leave because their creativity is being slowly suffocated by kindness: “We love your idea… let’s just run it through one more committee.”
The Edge of Chaos – Where Real Innovation Lives
There’s this concept in complexity science called the “edge of chaos.” It’s that narrow, uncomfortable zone between perfect order (where nothing new ever happens) and total chaos (where nothing coherent ever happens).
Healthy forests live there. Immune systems live there. Successful startups definitely live there.
Banks? Most are camped so deep in the order side that they need a passport to visit the edge.
But here’s the thing nobody wants to admit: you cannot regulate your way to the future. You can’t compliance your way to relevance. You can only explore your way there. And exploration requires friction, failure, and the occasional bruised ego.
Controlled Chaos: The Only Strategy That Actually Works
So what does deliberate, responsible disorder look like inside a 150-year-old bank?
It looks like small, ring-fenced teams with their own budget, their own tech stack when needed, and – crucially – their own failure quota.
- A team of 8–15 people max
- 12–18 month runway, no questions asked
- Direct reporting line to someone who can actually say yes
- Pre-approved compliance “sandbox” boundaries
- Permission to break things that aren’t customer funds or regulatory red lines
- Celebration of intelligent failure (yes, actually celebrate it)
I’ve seen this work. A Middle Eastern bank I advise stood up an internal venture unit in 2023 with exactly this model. Eighteen months later they have:
- A tokenized real-estate fund running on private blockchain
- An embedded finance partnership with the country’s largest e-commerce platform
- Two revenue-positive products the main bank still doesn’t fully understand (which is fine)
The parent bank didn’t lose control. It gained optionality.
Why Most Internal Innovation Labs Fail (And How to Not Be Most)
Here’s the dirty secret: 90% of bank “innovation labs” are theater. Glass walls, bean bags, a fintech event sponsorship, and zero actual power. They exist to make the annual report look modern, not to change the P&L.
Real internal ventures are different in four painful but necessary ways:
- They can kill sacred cows. If the venture proves the legacy product is obsolete, the legacy product actually gets sunset. Most labs aren’t allowed to discover that.
- They have real money. Not marketing budget. Real balance-sheet money they can deploy without 19 signatures.
- They can hire weirdos. The regulatory team has veto rights on risk, but not on personality. You need people who make compliance flinch – that’s how you know they’re pushing boundaries.
- Success means graduation or death. No perpetual lab sucking budget forever. Either the venture becomes a new business line or it gets shut down with dignity. Limbo is the enemy.
The Psychological Safety Paradox
Perhaps the most interesting aspect – and the hardest to implement – is cultural.
Banks reward perfection. Ventures reward speed and learning. Those are almost opposite operating systems.
I’ve found that the only way to square this circle is to make failure prestigious. Yes, prestigious. When a venture team kills their own project after burning $8 million because the market moved, that should be a promotion moment, not a career risk. Because they just saved the bank from burning $800 million on a dying idea.
The most dangerous experiment is the one you’re afraid to run.
The Coming Reinvention Wave
We’re already seeing the early winners.
JPMorgan’s Onyx. DBS Digibank. Goldman’s Marcus (before they half-killed it). Standard Chartered’s SC Ventures. These aren’t accidents. They’re institutions that figured out how to run two operating systems at once: one for today’s multi-trillion-dollar franchise and one for tomorrow’s.
The rest are quietly preparing their “we got disrupted” narrative for 2030.
Because here’s the uncomfortable truth: the disruptors aren’t coming from outside anymore. The neobanks peaked. The big tech platforms got scared of regulators. The next wave of billion-dollar financial innovation will either come from inside existing banks… or it will come from nowhere and catch them asleep.
Tokenization alone is going to rewrite the rules of asset ownership. AI will automate 80% of middle and back-office jobs in the next decade. Embedded finance will make the traditional bank account feel as archaic as a fax machine.
Banks that treat these as “watch items” will become museums.
Banks that treat them as experiments run by small, empowered, slightly scary teams will own the next century of finance.
Your Move
If you’re sitting in a bank reading this and thinking “sounds great but my organization would never…” – congratulations, you’ve just diagnosed the problem.
The question isn’t whether your bank can afford to run controlled chaos experiments.
The question is whether it can afford not to.
Because the alternative isn’t stability.
It’s slow, dignified extinction.
And nobody ever put that on a marble plaque in the lobby.
