Lancer 20 produits en parallèle, la méthode Revolut pour industrialiser l’innovation

Revolut has implemented a structured and scalable innovation process, diverging from traditional long and centralized development cycles. Over the past three years, Revolut has launched 27 product initiatives, with five becoming significant growth drivers. This article outlines their high-frequency innovation model.

Revolut's product development is not intuition-based. Each product launch is a calculated bet, executed by an autonomous team of approximately ten individuals. These teams operate independently with an annual experimentation budget of 2 to 3 million pounds.

Nik Storonsky, the CEO, states, "We can run 20 bets simultaneously without friction, as each team is independent." The goal is to test cost-effectively without overwhelming existing structures.

The framework relies on a straightforward rule: products must demonstrate measurable traction within the first three months. If growth metrics do not rise quickly, the team is not expanded. Conversely, if a bet is successful, additional resources are allocated to accelerate growth.

Storonsky explains, "We compare the revenue generated by each bet over the first four months. The top 30-40% earn the right to scale, transforming a team into a department." Recent successful bets include the eSIM offering and RevPoints, the integrated loyalty system.

Revolut has developed an internal interface to manage its portfolio of bets. Each initiative is tracked in Jira and aggregated in dashboards that centralize performance metrics such as gross revenue, growth rates, and adoption. The development stage (development, production, scaling) is also visualized.

This dashboard enables rapid decision-making by leadership without dependency on cumbersome committees. The decision-making process is data-driven and localized, with each product evaluated independently.

Assumed Success Rate and Optimized Yield

Of the 27 bets launched, about five have been clear successes, five to six have failed, and the remainder are in between. This yield aligns with a "venture studio" approach: few winning projects but with high growth potential. Storonsky notes, "If you look at overall cash generation, the portfolio is exponential."

The initial filter is not solely based on perceived potential. Revolut consults a committee of former product managers, and projects are launched even with mixed reviews.

Ability to Quickly Discontinue Non-Performing Products

The "Salary Advance" product case exemplifies this culture of rapid iteration. Designed as an alternative to consumer credit, the product allowed unlocking part of the salary before month-end but did not gain traction due to a complex adoption funnel.

Storonsky concludes, "For a lot of effort, we only had a few hundred users. We stopped it."

This "multi-bet" system makes Revolut an exception in the banking industry. Unlike neobanks that quietly iterate on a single product, Revolut builds a comprehensive service portfolio—credits, eSIM, loyalty points, trading, etc.—with a modular expansion logic.

The key lies in team independence, rapid testing, and selective discipline. Each product is judged on its intrinsic performance, and only the best receive resources for large-scale deployment.

Bets Launched in 3 Years: 27 Concurrent Bets: 20+ Average Budget per Bet: 2 to 3 million pounds Clear Success Rate: 5 out of 27 Scaling Mode: Team progression to department Selection Criteria: Traction within months 1-3, gross margin, adoption

D’après FrenchWeb.