Competition, Intensity & Building a $115 Billion Global Bank | Nik Storonsky, Revolut

Episode Overview

• Host David Senra (of the David Senra and Founders podcasts) interviews Nik Storonsky, former financial trader and founder and CEO of Revolut and QuantumLight, on his mission to build the world's first truly global digital bank and become the #1 bank in at least 100 countries.
• The conversation explores how Storonsky's lifelong hyper-competitiveness—dating back to age 6 in swimming, boxing, and competitive olympiads—shapes Revolut's relentless operating system, contrasting founder-led intensity against traditional banks run by short-term professional CEOs.
• Storonsky breaks down his algorithmic approach to product velocity, AI integration, and talent management, detailing how small, independent teams operate on strict metrics and why he eliminates middle management and fires "average" performers.
• The episode covers Storonsky's hard-won lessons on banking regulation, his removal of traditional venture capitalists from his board, and the creation of his own machine-learning-driven VC fund, QuantumLight.
• Finally, the discussion reveals how Revolut survived a 70% COVID revenue drop to master marketing arbitrage, develop its own transactional foundational AI model (Pragma), and prepare to challenge US banking giants like JPMorgan while planning future deep-tech moonshots.

Key Concepts

• The Global Banking Opportunity and Founder Advantage: While there are tens of thousands of banks globally, individual countries typically have 10, 20, 30, 40, or 50 fragmented local banks. Revolut, which Storonsky started 12 years ago at age 29 with a simple card for spending abroad at interbank rates, now has tens of millions of customers across credit, wealth/trading, crypto, business accounts, and non-financial services like booking stays and eSIMs. It is currently live in "30, probably 38 countries" (with Storonsky hedging slightly on the exact count), holds banking licenses in the majority of them, and expands into 10 to 15 new markets every 1 to 2 years. Storonsky argues traditional banks are easy to beat because they are run by non-founder professional CEOs who stay for 3 to 4 years at best to collect bonuses, play defense, lack the technical skills to migrate off legacy third-party core systems, and view themselves as separate from the company. In contrast, Storonsky views himself and Revolut as one entity with 100% alignment, dedicating 99.5% of his time to building the business and focusing on beating himself at "almost 100% every day." When asked if Revolut's US expansion is coming for JPMorgan CEO Jamie Dimon's lunch, Storonsky replies he is coming for his "dinner"—targeting retail, enterprise, and manual, expensive investment banking. Revolut's advantage over JPMorgan includes global exposure (instant cross-border transfers and multi-country accounts), a better and cheaper interface, a subscription and transaction revenue model (versus a credit focus), and a single, configurable global software ledger (whereas legacy banks grow by acquisition and take months to pull data across fragmented systems).
• The Regulatory Paradox and Market Penetration: Storonsky identifies not getting a full banking license on day one as a major mistake. Revolut initially pursued e-money and payment licenses because they took 1 year, whereas a banking license took 3 to 4 years. However, by the time Revolut applied for a UK banking license, it had 7 or 8 million UK customers—unprecedented for an applicant. Because regulators are far more risk-averse toward companies with 5 to 10 million customers than startups with zero customers, UK approval took almost 5 years. Furthermore, Storonsky notes that although regulators view him as an excessive risk-taker due to his growth mentality, he is the least likely to take stupid risks because he is 100% motivated for the company to survive. He argues US fintechs like Chime, Cash App, and Simple made a big mistake 10 to 12 years ago by skipping banking licenses; without one, you lack customer trust for salaries and wealth, relegating you to a 4% niche (1% to 2% early adopters plus 2% friends and family) instead of capturing 40% to 70% of the prime market. Today, Revolut has 80% market penetration in Ireland (4 out of 5 adults, driven by a large expat population and only 4 bad local banks), about 20% in the UK, 20% in France, 20% in Spain, and 30% to 40% in other markets. Surviving Europe made Revolut much stronger: unlike the US (with 1% to 1.5% interchange fees and at-will employment), Europe caps interchange fees at 20 basis points, caps credit card rates, requires local branches per EU regulation to issue local accounts, spans multiple languages, and has tough employment laws. Revolut endured negative gross margins for its first 3, 4, or 5 years before optimizing to up to an 80% positive gross margin.
• Organizational Speed, the "Bets" System, and AI Integration: When Revolut reached 10, 20, or 30 people, speed slowed down, prompting Storonsky to restructure into small, independent teams working in parallel on the same platform. Over the last 10 years, Revolut has launched roughly 50 staged "bets." Historically, a committee approved a bet and assigned a separated team of up to 10 people (typically 3 to 4 backend developers, 2 to 4 frontend developers, 1 product owner, 1 designer, and sometimes a data scientist or operational manager) to build it within 9 to 18 months. Storonsky explicitly does not trust his own product intuition because he was wrong so many times historically; instead, bets are judged algorithmically. The top 10 bets grow 10% to 20% month-on-month and generate 1 million, 2 million, or 3 million in Gross Profit (GP) each month consecutively. Bets in the bottom 25% cohort within 18 months are sunsetted, while the top 25% are scaled from 1 team to 2, 3, 4, or 5 teams, then a department, and a major P&L line. Today, AI is shrinking these 10-person teams to 2 or 3 people at best. Revolut has automated 90% of its customer support team (formerly 5,000+ people), increased engineering speed by 30%, deployed LLMs for translation and cross-selling, and is building a "full-stack AI employee" capable of using Slack, email, back-office tools, and joining calls. While Senra notes Alexandr Wang has ~200 direct reports and Andrew Carnegie ran his company for 30 years obsessing over 5% improvements in 1% of his business, Storonsky maintains 50, 51, or 52 direct reports and frequently skips levels to speak with individual contributors—which recently allowed him to spot a design team feeding Revolut's code-level design system into an LLM to generate live simulator flows 3 to 4 times faster than static Figma frames.
• Talent Tiers, Capital Discipline, and the 996 Work Ethic: Storonsky eliminates middle managers who merely pass numbers between individual contributors and stakeholders without adding innovation or changing direction. He categorizes talent into four tiers: Brilliant ("self-guided missiles" who self-identify and reach goals without being told), Good (reach goals once told, without weekly iteration), Average (require weekly management iterations to hit goals, which wastes time, so he fires them), and Below average (fail to hit goals even with management time). He argues executives fail to fire average people because they prioritize empathy and being likable (#2, #3, or #4) over winning, whereas Storonsky has been "extremely pissed off" by being #2 since age 6. He also criticizes US and California startups for raising too much money at ridiculous valuations and solving problems with money rather than brainpower—such as hitting $1M ARR and immediately hiring an expensive VP of Sales and 50 salespeople instead of building a cheap, scalable system. He views China—not the US or Europe—as the truly competitive market due to great technical education and a "996" work ethic (9 am to 9 pm, or 9 to 10, 6 days a week). Observing his first co-working space of 140 companies, only 3 survived within 5 years (all 3 worked weekends) and only 2 survived within 10 years, leading him and most of his direct reports to work a 996-style schedule.
• The VC "Cheerleader" Problem and QuantumLight: Echoing a point Senra cites from Travis Kalanick—that over 90% of VCs can't avoid messing things up and only 1% are helpful—Storonsky compares VCs to "cheerleaders" on the sidelines who mistakenly believe providing capital entitles them to tell professional players on the field how to play, despite VCs operating an easy 2/20 business model with LP money and no operational experience. Following VC advice in his first 5 years, Storonsky hired 50+ "experienced executives" and kept only 2 within a year. He attributes this failure to a "prime age" effect (shifting priorities) and "negative selection bias" (truly elite executives have already made fortunes, retired, or run their own companies). After a prominent investor formed a coalition to try to force him to fire his best person for being "too hardcore," Storonsky fought back and used subsequent funding rounds to remove all VCs from Revolut's board (now at zero). Later, frustrated by a big-name investor who invested at a $33 billion valuation during/after COVID and blocked shareholder agreement changes, Storonsky founded his own VC fund, QuantumLight, around 2022 or 2023 (hedging on the exact year and whether it will reach its goal of becoming the BlackRock of private markets). QuantumLight uses machine learning models—not human judgment—to statistically select 30 to 40 companies over 2 to 3 years with equal check sizes. Fund 1 (Series B) is $250 million (with Storonsky disclosing he invested 20% to 25% of his own money and the rest from other founders) and is fully invested, producing a 56% IRR. Fund 2 (Series C) is $500 million (including institutional LPs) and ranks in the top 5% of funds globally. Notably, QuantumLight's ML model found that startups backed by VC partners over age 40 underperform, while VC partners aged 30 to 40 perform best (though Storonsky hedges that age is not the only variable, just what to look at statistically).
• Marketing Arbitrage and Deep Tech Moonshots: Revolut spent $0 on marketing for its first 6 to 7 years (100% organic) until COVID hit in 2020/2021 and travel-card reliant revenues dropped 70%. To start marketing, Storonsky targeted the industry with the largest budgets ($1B+) and lowest margin of error—mobile gaming—and hired a gaming marketing lead (after firing his first marketing hire). To overcome the trust deficit digital banks face against branch-heavy banks like HSBC, Lloyds, or JPMorgan, Revolut arbitraged digital performance marketing by purchasing cheaper physical-world impressions: buying ad space in ~50% of all European airports and opening an Apple Store-style "Rev Store" in Barcelona where up to 1 million people pass annually. Looking ahead, Storonsky compares his ambition to Meta: he wants to use Revolut's massive cash flow to fund capital-intensive deep-tech moonshots (such as data centers and foundational models) outside financial services—hedging that he hopes to have the resources and time in 3, 4, or 5 years. Internally, Revolut has already built Pragma, its own foundational transformer model trained on Revolut's 30 million daily transactions (billions total) rather than words, powering anti-fraud, credit, and investment modeling.
• Disclosed Business Ties and Sponsor Claims: Host David Senra mentions his partner "Rob" and discloses that he runs his own business on presenting sponsor Ramp, stating that the median company on Ramp cuts expenses by 5% and grows revenue by 16%. Senra also shares sponsor claims for Deel (trusted by over 40,000 companies) and AppLovin (which sponsored the video production and advertising, connects businesses with over 1 billion mobile gaming customers, averages 35-second views on full-screen video ads, and will print ~$6 billion in profit this year).

Quotes

• At 0:00:40 - "Like if I choose to play a game, I need to be number one. And then I'm extremely pissed, pissed off if I am number two. So I always had this character." - Storonsky explaining the lifelong hyper-competitiveness driving Revolut's ambition to be the #1 bank in 100 countries.
• At 0:05:04 - "I mean, to be honest, I don't really look at competition. I look at kind of myself, right, and then what would I do to beat up myself, right, to outcompete, to outsmart somehow. And the philosophy is very simple: so every day you need to be like almost 100% every day." - Storonsky describing his internal benchmark for continuous improvement rather than fixating on external banking rivals.
• At 0:07:22 - "Well, professional CEOs, ultimately their goal is not to improve organization, right? Their ultimate goal is to protect their job. So they, by definition, they make suboptimal decisions for, for the company." - Storonsky contrasting founder alignment with the defensive mindset of traditional bank executives.
• At 0:09:19 - "Because for them it's much easier to, to give a, a license to a company with a zero customers, because maybe they will have zero customers in three, four, five years' time, right? If you give a banking license to a company with five, ten million customers and something goes wrong, right, then the regulator will be to be blamed." - Storonsky explaining why delaying a full banking license application until Revolut had millions of users was a major strategic mistake.
• At 0:19:52 - "Metrics and KPIs, they actually kills politics within organization because, look, everything is transparent, right? You as a individual contributor or you as a manager, did you achieve your metrics? Did you achieve your goal, right? Why you haven't achieved it? So it doesn't matter what you say, I mean, there are objective numbers." - Storonsky detailing how quantitative transparency eliminates internal corporate politics.
• At 0:22:16 - "Even myself, I kind of, you know, I don't really trust my intuition in terms of which products to build, which products not to build, because historically I was wrong so many times. And I think only approach which like completely algorithmic, right—top bets you continue to fund, bottom bets you just cut like within 18 months." - Storonsky expressing humility about his own product intuition and explaining Revolut's algorithmic "bets" framework.
• At 0:24:01 - "I think the team should not be 10 people anymore, it should be two people, three people at best, because now one developer can do frontend, backend, can even do, you know, data science as well. So it should be rethought." - Storonsky explaining how AI tools are fundamentally compressing software product team sizes.
• At 0:31:31 - "So I think if you're much more weighted in terms of winning, right, your empathy to other people is a second place, and then you, you will do right decisions even if they are tough decisions." - Storonsky explaining why executives must prioritize winning over being likable when firing average performers.
• At 0:33:48 - "So I think there is a definite culture of trying to solve problems with money rather than with your brains." - Storonsky critiquing overfunded US startups that hire large sales teams prematurely instead of engineering scalable systems.
• At 0:35:35 - "When I started, I started in a co-working space with 140 companies. Within five years' time, only three companies survived; they all worked on weekends. Within 10 years, only two companies survived out of 140. So I think working on weekends is a must if you want to survive, purely statistically." - Storonsky sharing the observational data behind his commitment to a 996-style work schedule.
• At 0:45:08 - "Because if something goes wrong, right, you effectively have cheerleaders, you know, if you look at the match, and then they try to play this match, you know, instead of professional players, and that's, you know, always disaster." - Storonsky illustrating the danger of allowing non-operating venture capitalists to dictate company decisions from the boardroom.

Takeaways

• Allocate product resources algorithmically rather than by intuition: Structure new initiatives as staged 9-to-18-month "bets" built by small, autonomous teams (2 to 3 versatile people in the AI era). Measure success strictly by month-over-month gross profit growth against historical cohorts—automatically cutting the bottom 25% of projects within 18 months without emotion and aggressively scaling the top 25%.
• Hire for career velocity rather than legacy executive resumes: Avoid the negative selection bias of hiring "experienced executives" who may be past their prime. Instead, recruit ambitious talent aged 25 to 30 who rank in the top 1% of career progression among their peers, and retain only "brilliant" (self-guided) or "good" employees who execute without requiring weekly management iterations.
• Flatten hierarchies and skip levels to scale technical innovation: Eliminate middle managers who merely relay numbers without adding innovation. Maintain a wide span of control and regularly skip management levels to speak directly with individual designers, developers, and product owners—spotted innovations (like replacing static Figma workflows with code-level LLM simulator prompts) can then be mandated across the entire organization.
• Tackle the hardest regulatory hurdles on day one: If operating in a regulated industry like finance, apply for full licensing immediately while your customer count and systemic risk are zero, rather than relying on interim licenses that cap prime-market trust and trigger multi-year regulatory delays once you have millions of users.
• Protect founder control and board governance: Do not allow investors without operational experience to dictate personnel or operational strategy; structure financing rounds to maintain board control and evaluate capital partners quantitatively based on the individual partner's active track record.
• Arbitrage physical and digital marketing channels using pure math: Recruit marketing leaders from high-budget, zero-margin-of-error industries like mobile gaming, and look for mathematical arbitrage where high-traffic physical spaces (such as airport corridors and busy retail streets) deliver lower cost-per-impression and higher brand trust than digital ads.