Antoine Le Nel on Revolut’s Growth Engine and Marketing Measurement

Harry Stebbings asks Revolut chief growth and marketing officer Antoine Le Nel what made Candy Crush successful at King. Le Nel credits a high volume of experiments and rapid releases: the company launched games frequently, knowing not all would succeed. The hard skill, he says, is cutting projects when data shows they are not working.

King became the most data-driven organization he had seen, but he thinks it eventually went too far, sometimes reviewing numbers more than the game itself. At Revolut, in his view, product experience carries more weight and data is an input. He sees product work as more artistic and growth work as more analytical, an emphasis rather than a strict division.

Le Nel calls performance marketing a competitive advantage when mastered. The underlying auction for impressions remains a bidding war, he says, even as attribution has become harder and measurement frameworks need to evolve. These are his professional judgments, not evidence that any one channel will work for every company.

At King, he recalls roughly 1,000 A/B tests running in parallel as a mature business optimized details. Revolut had far fewer tests when he joined. He argues a strong product manager can sometimes know the right direction without a test, and that testing may slow launches when a large improvement is already clear; he expects more testing as the company matures.

He describes Revolut as founder-led, with the founders setting a consistent tempo. Asked how growth can keep accelerating, he says there is no single campaign responsible: many elements of a growth engine compound and are optimized week by week. This is his description of the company’s performance and strategy, not independently verified growth data.

Revolut organizes teams to be as self-sufficient as possible, with a flat structure and quarterly cycles, he says. Autonomy reduces dependencies and increases speed, though teams still need coordination. Campaigns that fail are cut quickly: he describes stopping generic creative after two weeks rather than waiting for a planned campaign to finish.

A growth engine connects the upper, middle and lower parts of the funnel. Le Nel likens the conversion layer to a net that captures demand generated by awareness and consideration. The balance changes with market maturity; in a new market he emphasizes conversion, while a mature one can support more upper-funnel work.

Measurement is the hardest part, he says: teams naturally favor activities they can measure, so better attribution can open opportunities. He recalls King developing ways to connect television advertising to downloads earlier than competitors. Changes to iOS attribution also require more sophisticated measurement rather than assuming observed downloads capture the full effect.

He would build the growth engine from the beginning, including virality, not just paid marketing. His preferred division is for product teams to build the experience and configurable mechanisms, and growth teams to operate and measure them. He warns against fixed beliefs about particular channels: the evidence should determine what earns further investment.

At King he was skeptical of upper-funnel brand marketing. At Revolut, he says, testing changed his mind: brand activity can improve conversion. As an example, advertising at airports is paired with vending machines for Revolut cards, allowing the company to link some account creation to the activity. He says attribution helps, but the interview does not independently validate the return or isolate every brand effect.

Some celebrity or influencer campaigns failed, particularly when the person was not effective on social media, he says. Le Nel describes mid-funnel activity as combining reach and brand consideration with measurable response, while lower-funnel work includes performance marketing, affiliates and referrals. He prefers specialists in each area, with generalists often suited to management.

Launching a new country starts the funnel again, though awareness may spill over from nearby markets. Stebbings asks whether customer acquisition cost falls as brand and word of mouth grow or rises as the addressable audience saturates. Le Nel says both forces operate, and the outcome depends on competition and the individual market.

Le Nel says Revolut focuses its discussion on return on investment rather than setting customer-acquisition-cost targets alone. His reasoning is that the cheapest users may be less valuable: paying more for a cohort can be worthwhile if its subsequent contribution is better. Stebbings points out that the higher return may take time to appear; Le Nel says models provide early indications, which remain estimates.

He believes some cohort models can transfer across countries, but adjusts them for the product sold. Stebbings notes that payment practices can differ substantially between places such as London and Brazil. Le Nel accepts that distinction and says the model must reflect the relevant product and context.

Revolut wants eventually to operate everywhere and prioritizes markets partly by ease and speed of entry, he says. His claim that localization is often overrated is deliberately provocative: he recalls similar creative working in many countries at King, with Japan an exception, and says Revolut used a similar strategy in Romania and Norway. This is his experience, not proof that regulation, language or culture never require local adaptation.

Stebbings asks about the share of users who make Revolut their primary account. Le Nel says the objective is to become a primary bank for more people, but not to force a deadline: a user might start with travel, then try investing or savings, and gradually move more activity. Stebbings describes his own similar progression.

Asked about a crypto-driven spike, Le Nel says he welcomes demand and views crypto as part of the financial product range. His company offers such products, so the view has a commercial interest; a spike in activity does not by itself establish sustained customer value or suitability for everyone.

He tracks three things: user volume, return and payback from acquiring them, and progress toward being a primary account. He says payback varies by market but the framework has been relatively stable. He also cautions that an effective growth engine must keep evolving rather than assume today’s playbook will work forever.

Le Nel says Revolut sets ambitious targets that are not always reached. Stebbings asks whether perpetually missing goals damages morale; Le Nel argues that achieving roughly 70–80% of a demanding target can still be a good outcome. Their exchange leaves open how teams experience those targets in practice.

The growth organization has roughly 20–25 teams across channels and local markets, he says. He runs weekly business reviews with them, focused on solving problems rather than preparing presentations. Short cycles, in his account, allow frequent feedback and course correction.

Le Nel says he has 20–25 half-hour reviews each week and relies on recurring meetings rather than a changing schedule of ad hoc ones. Some meetings have 10 to 40 participants. He believes the meetings also transmit culture in a remote organization, even to people who do not speak during every session.

He favors transparent, decisive yes-or-no calls and admits they can produce wrong approvals or rejections. Stebbings contrasts this with a more discussion-heavy approach and asks whether speed harms creativity. Le Nel believes shorter decisions allow more iterations and therefore more creative output, though the trade-off is less discussion before each decision.

On the US, he says the market is more credit-oriented than much of Europe and observes that local fintech penetration differs across countries. He suggests some developing markets leapfrogged legacy infrastructure, as with mobile telecommunications. This is his interpretation of market differences, not a demonstrated causal explanation.

Le Nel works remotely from Barcelona and acknowledges the loss of human contact, particularly for people early in their careers. The team meets in person about once a year. He argues remote work makes short meetings easier and can give quieter people a better opportunity to speak through features such as raising a virtual hand.

Stebbings pushes back that people need to learn to interrupt politely and that valuable post-meeting conversations can disappear remotely. Le Nel says he follows business reviews with individual feedback sessions and expects team leads to do the same. The two disagree on whether that reproduces the informal learning available in person.

For hiring, he uses case studies to see whether candidates can break a problem into manageable parts. He argues growth teams often hire generalists who do many things adequately rather than specialists who go deep. He questions vague “growth manager” titles and asks whether the actual role concerns referrals, paid marketing, brand, pricing or another function.

Asked whether growth teams should pursue moonshots or incremental improvements, he strongly favors the latter. Small improvements in creative, bidding and funnel conversion can add up in an auction, he argues. His athlete analogy illustrates marginal gains; it does not establish that all growth channels are literally winner-take-all.

Revolut still has work to do on brand positioning, Le Nel says. He sees brand first as a product experience and later as something marketing can amplify. Stebbings argues both product design and advertising contributed to Apple’s brand; Le Nel agrees the balance changes as a company matures.

Le Nel warns early founders against treating weak awareness as the only explanation for poor growth. He believes product quality and adoption should come first, with broad brand marketing becoming more useful when a company needs to reach new segments. This is his suggested sequence, not proof that strong products always attract customers unaided.

He says Revolut’s crypto exchange had not yet grown as much as expected, though it was performing, while airport lounges and premium subscriptions were working in some markets. Stebbings questions the value of a roughly £600 annual plan; Le Nel points to benefits and loyalty features. These are company-side product claims, not a recommendation to purchase financial services.

Le Nel praises CEO Nik Storonsky’s resilience and ability to move from a broad view into a specific problem. He has adopted weekly problem-solving reviews rather than meetings devoted mainly to presenting successes. The aim is to identify obstacles and solve them together.

In the quick-fire round, he says performance marketing remains important while direct sharing has become harder. His advice to a new growth leader is to build an analytical framework before spending. He criticizes platforms that, in his view, invoke privacy while encouraging unmeasured brand spending; that is his opinion, not an assessment of particular privacy protections.

He would like Revolut to offer mortgages and considers the category important to customer perception, while saying the company is working on it. He admires Nubank’s depth in Brazil and its growth in Mexico. These are aspirations and competitive observations from a Revolut executive, not a claim that a mortgage product was already available.