Val Scholz on Revolut’s Referral Growth, Hiring and Scaling

Former Revolut growth leader Val Scholz argues that startups can find markets with less direct competition and focus on a product substantially better than the existing alternative. He points to Revolut’s early expansion in Ireland, Romania and Poland as examples. These are his retrospective competitive assessments.

Scholz tells Harry Stebbings that he began by building websites and learned search engine optimization from a friend. At MindMeister he worked on sharing mind maps on external sites to generate traffic, an early experience of what would later be called growth.

At Revolut, he says the main lesson was to do one thing exceptionally well rather than pursue many at once. He claims roughly 90% of consumer customers acquired during the period came through referrals. The interview does not specify the measurement window or independently verify the figure.

The team sought to make signup and a first payment possible within five minutes so a recommendation could turn into use during a conversation. Scholz argues that referral speed matters as well as the number of people invited: three referrals over three minutes and three over three years produce different growth rates.

He estimates first-payment activation rose from about 30–35% in the early days to around 90% when he left. After emailing 4,000 users, he says he found that many did not yet have a use case because they saw Revolut as a travel card. The exact cohort, observation window and independent basis for the comparison are not supplied.

The team changed positioning and addressed friction, including identity verification and country-specific ways to fund accounts. Scholz says higher referral rewards attracted more users but had to be balanced against acquisition cost. He recalls an approximate £9 cost per acquisition in one period; Stebbings points out that cost per activated user can be higher when only some signups make a payment.

Scholz says early referrals offered a free card rather than cash because the company needed funds for product, engineering and payment operations. He recalls that Revolut’s original foreign-exchange use case later supported additional products such as savings and investments. These are his accounts of strategy, not an audited cost breakdown.

He believes additional products can make growth easier by giving customers more reasons to use a service and allowing cross-selling. He says a paid premium offering gained traction and recalls strong demand for a metal card, including roughly 10,000 preorders. The figures and reasons for demand are his retrospective account.

The team tried many experiments, including handing out cards at technology-company offices, though long-term effects could be hard to measure. When a channel such as YouTube influencers appeared to work, he says the team expanded it quickly. He recommends estimating views, click-through, conversion and acceptable acquisition cost, then testing several niches; his example assumptions are not guaranteed rates.

Because lifetime value is uncertain early, Scholz recommends comparing monthly cohorts and watching whether retention stabilizes. He recalls early payment activity settling around 20% after several months and later improving as products for everyday banking were added. The interview does not define the full cohort methodology or prove which changes caused the improvement.

He contrasts labor-intensive traditional marketing with software-driven loops such as referrals, paid incentives and user-generated content. He says Revolut did not rely on paid marketing in his period because spending stops producing traffic when switched off. This is his model and historical account, not proof that all paid marketing is ineffective.

For content acquisition he describes two approaches: user-generated content and internally produced content organized around search topics and landing pages. An internal team can seed supply before a user community creates it. He cites Canva as an example, but provides no independent measurement of its search traffic.

Scholz advises doubling down on a channel that works before spreading limited resources too widely. Running many channels early can make attribution noisier and learning slower, although he says some experimentation elsewhere can still be useful. The appropriate balance depends on the company and its data.

He prefers hiring a head of growth after demonstrating product-market fit, because hiring one creates pressure to show immediate growth. He argues that traffic can temporarily support a business even with weak retention, but that a durable product requires a distinctive insight and genuine demand. His examples and causal interpretation are not a controlled study.

For a first growth leader, he favors an ambitious person who can grow with the company. He describes the role as stressful because performance is frequently measured. He believes hiring mistakes arise when interviewers do not probe candidates’ actual contribution or whether their motivations match the role.

Scholz says Revolut looked for driven people who could learn quickly and believed skills could be taught more readily than attitude. He recommends being explicit about expectations, including evenings or weekends, so candidates can decide whether the environment suits them. His personal preference is not evidence that long hours are necessary for all workers or teams.

He describes sourcing as a funnel: identify relevant experience, contact candidates and have hiring managers screen them. He is skeptical of résumés that emphasize budget or team size without explaining impact. He used elite universities and rapid promotion as rough signals of learning speed; these proxies can miss capable candidates and are his own hiring approach.

For interviews, the team gave candidates difficult past problems using altered data and looked for a structured approach rather than a complete solution. Scholz says a repeatable method for identifying causes and testing hypotheses mattered more than a memorized answer.

He says he often formed an early impression of a hire, tried to improve the situation for three or four weeks, and sometimes moved on if progress was insufficient. He looks for whether people respond to feedback and learn. This is a personal management heuristic, not a reliable universal probation period.

Asked about culture, Scholz says Revolut was a demanding environment designed around rapid performance and that a person who does not fit it may succeed elsewhere. He compares it with elite sports teams. His positive account is a former manager’s perspective, not a survey of employees’ experiences.

He recalls coaching a creative designer who was repeatedly refining one screen while the team needed a complete user journey. Scholz increased check-ins to four a day for several weeks, then reduced them after the work improved. He uses the story to argue for time and feedback, though it is one anecdote rather than evidence of a general method.

He favors frequent feedback cycles, suggesting weekly reviews as a balance between learning and having time to work. He recalls founder Nik Storonsky attending many team meetings and says access to direct feedback helped employees develop. The interview does not measure the effect of those meetings.

Asked whether the founder led through fear, Scholz describes a truth-seeking culture where he says data could change Storonsky’s mind. He acknowledges some people may have felt pressure or fear of being wrong. This is his interpretation and does not settle how others experienced the workplace.

Rapid growth strained every system, Scholz says: headcount, communication, databases and support all had to scale. He recalls a Romania referral launch rising from roughly 2,000 to 64,000 daily signups in three days, after which the team paused it. He says identity checks and support waits lengthened sharply; the numbers are his recollections and their precise definitions are not supplied.

He says growth tactics varied less between countries in the earlier period, while later reaching older users could call for new channels. Revolut positioned itself as an alternative to a traditional bank rather than focusing on the label of primary account. These are his historical product and marketing views.

On data infrastructure, Scholz criticizes reliance on isolated event analytics when payment, campaign and product data need to be connected. He describes consolidating sources in a data warehouse and joining them by user ID to examine cohorts and economics. This is his architecture preference, not evidence that every organization should abandon event tracking.

He recalls a quick referral campaign offering free cards that produced about 180,000 signups in a week versus a prior pace he recalls near 10,000 per week. His least successful experiment tried to recruit users through payments to non-users; he suspects recipients may have distrusted unexpected payment messages. The figures and explanation are not independently verified.

In quick fire, he urges founders to use their own products and warns that relying on paid acquisition from the start can make product-led growth difficult. He predicts AI will initially increase automated outreach and noise, making genuine value and ways to reach people important. These are practical views and predictions.

He considers Revolut a broader financial ecosystem than Monzo and praises Storonsky’s analytical and creative thinking. He describes the difficult expectation of high-quality work at maximum speed. His product comparison reflects his former role at Revolut and is not an independent feature assessment.

Scholz speculates about a future where physical robots change how people interact with digital systems. He also discusses Kittl, where he works, as an underdog facing larger design competitors. He argues smaller firms can enter segments incumbents overlook; his strategic view involves his current employer.

He praises Nubank’s reported retention and installment-payment features, describing a discount for paying early, and cites rough user and profit figures. His explanation of Brazilian rules, economics and company metrics is not sourced in the interview and should not be treated as a verified legal or financial account. The conversation ends with thanks.