The Twenty Minute VCAntoine Le Nel, CGO @Revolut: How Revolut Launch and Grow Products & Why CAC is a BS Metric | E1216
CHAPTERS
- 0:00 – 0:44
Revolut’s growth philosophy: ditch CAC, trust product intuition
Antoine sets the tone with contrarian takes: strong product managers shouldn’t need constant A/B tests, and Revolut doesn’t manage growth by CAC. Instead, the team focuses on ROI and cohort quality, framing paid growth as a competitive bidding environment.
- •A/B tests can be a crutch; good PMs should know what’s right
- •Revolut avoids CAC as a primary metric and focuses on ROI
- •Chasing low CAC tends to attract lower-quality cohorts
- •Growth is fundamentally a bidding war; winning requires outbidding with better economics
- 0:44 – 1:38
Candy Crush at King: volume of experiments and rapid shipping
Antoine explains King’s core success factors behind Candy Crush: relentless testing and an unusually fast cadence of launches. The organization treated growth as a volume game—ship many bets, learn quickly, and let winners emerge.
- •Extensive testing as a systematic advantage
- •High launch cadence (roughly quarterly game releases)
- •Not every release succeeds; portfolio approach matters
- •Speed and repetition compound learning
- 1:38 – 2:16
Knowing when to kill projects: agility equals fast cuts
The discussion turns to the hardest operational skill: stopping initiatives when data and traction don’t support them. Antoine highlights how organizations tend to become less decisive over time, and that the speed of killing projects is a proxy for true agility.
- •Early King cut projects quickly; later it hesitated
- •More ‘second chances’ often signal declining agility
- •Decision speed matters as much as decision quality
- •Learning loops depend on fast termination of weak bets
- 2:16 – 3:52
When data goes too far: King’s over-optimization vs Revolut’s product focus
Antoine contrasts King’s extreme data-driven culture with Revolut’s more product-driven approach. At King, metrics could become the sole decision-maker, sometimes at the expense of user experience and product quality.
- •King: data-driven to the point of ignoring the product experience
- •Revolut: data is an input, not the decision-maker
- •Product quality and UX can be underweighted in pure-analytics cultures
- •Product leans ‘art’; growth leans ‘science’
- 3:52 – 5:05
Performance marketing realities: competitive advantage, attribution loss, and reinvention
Antoine outlines why performance marketing mastery still creates outsized advantage, even as attribution becomes harder. While tooling and privacy changes add complexity, the underlying market dynamic remains the same: auctions and bidding wars.
- •Performance marketing is deep and difficult to master
- •Core mechanic hasn’t changed: impressions go to the highest bidder
- •Attribution loss forces better measurement frameworks
- •Teams must ‘reinvent’ measurement regularly to stay competitive
- 5:05 – 7:48
King vs Revolut operating systems: A/B testing, 10x bets, and time-to-market
Antoine describes the stark difference in maturity and process: King ran massive parallel A/B testing; Revolut often ships faster without it. Revolut optimizes for big step-changes and speed, arguing that you don’t need an experiment to see a 10x move.
- •King ran ~1,000 A/B tests in parallel at maturity
- •Revolut initially did little A/B testing; prioritized shipping
- •Three reasons to avoid A/B testing: PM clarity, 10x changes, faster time-to-market
- •Revolut is earlier in lifecycle: creating before optimizing
- 7:48 – 10:21
Why Revolut accelerates: the compounding growth engine and a flat org design
Antoine explains Revolut’s growth as steady, not spike-driven, because many components compound together. Organizationally, speed comes from flat structure, high autonomy, and minimizing interdependencies—many ‘startup-like’ teams moving in parallel.
- •Growth is not driven by one campaign ‘coup’ but many compounding parts
- •Acceleration increases as the engine gets optimized week by week
- •Flat org + self-sufficient teams increase velocity
- •Autonomy and limited interdependencies enable parallel execution
- 10:21 – 11:19
Cut fast, learn fast: stopping bad campaigns early
The team applies the same decisiveness to marketing: if a campaign underperforms, they stop it immediately rather than waiting for a planned end-date. Antoine gives examples of generic creative failing and being cut within weeks.
- •Poor performance triggers immediate shutdown (not end-of-quarter reviews)
- •Generic creatives can fail quickly in-market
- •Short feedback cycles reduce waste
- •Bias toward action over sunk-cost thinking
- 11:19 – 15:34
What a ‘growth engine’ actually is: full-funnel synchronization and measurement moats
Antoine defines a growth engine as synchronized upper-, mid-, and lower-funnel components that feed conversions reliably. The hardest part is measurement—companies win by measuring what competitors can’t, unlocking channels others avoid.
- •Engine = upper/mid/lower funnel working in sync
- •Lower funnel ‘fishing net’ captures what upper funnel generates
- •Measurement framework is the main bottleneck and advantage
- •Examples: TV attribution at King; post-privacy iOS measurement challenges
- •Start building growth engine from day one, including virality
- 15:34 – 20:01
Playbooks without ‘belief’: adapting tactics, upper funnel, and attribution hacks
Antoine argues growth engines fail when leaders operate on faith in channels rather than evidence. He shares how his King playbook changed at Revolut—especially embracing upper-funnel brand spend—supported by creative attribution approaches like airport vending-machine card pickups.
- •Avoid channel ‘belief systems’; treat everything as testable
- •Revolut shifted toward more upper-funnel brand marketing over time
- •Brand can improve conversion metrics when measured properly
- •Attribution workaround: airport ads + vending machines to link to sign-ups
- •Flops: certain influencers and celebrity campaigns that didn’t convert
- 20:01 – 29:06
Funnel specialization and international scaling: markets, CAC vs ROI, and localization skepticism
Antoine breaks down who owns each part of the funnel and emphasizes specialist operators over generalists. He then covers market expansion, arguing CAC is misleading; Revolut optimizes for ROI and cohort quality, and he claims localization benefits are often not worth the effort.
- •Upper/mid/lower funnel definitions and why they differ
- •Teams staffed with specialists; generalists mainly for management
- •Each new market restarts the funnel, but brand awareness can carry over
- •CAC can rise or fall; depends on brand strength vs saturation/competition
- •Revolut ignores CAC targets and optimizes ROI/payback and cohort quality
- •Localization is ‘overrated’; global creatives often work (Japan as exception)
- 29:06 – 31:54
Primary account is a journey: ‘snack’ adoption, product expansion, and crypto spikes
Antoine reframes Revolut’s path to becoming a primary bank: users adopt in small bites—travel, trading, savings—then consolidate over months or years. He also defends crypto as a legitimate financial product and dismisses criticism of growth ‘spikes.’
- •Primary-account conversion happens gradually, not via one switching moment
- •Revolut encourages trial and incremental usage expansion
- •Different user journeys take different time horizons
- •Crypto is part of modern finance; spikes are not inherently bad
- •Goal: be a one-stop shop for financial products
- 31:54 – 40:18
North Star metrics and operating cadence: weekly cycles, standardized reviews, transparency
Antoine shares Revolut’s core success metrics: user volume, ROI/payback, and progression toward primary usage. He then details the growth team’s meeting system—highly standardized weekly problem-solving reviews across many teams to maintain speed, alignment, and fast course correction.
- •Three metrics: user volume, ROI/payback months, primary-user progression
- •Payback is stable and standardized market-to-market
- •Hard part of the role: evolving the engine before it decays
- •20–25 teams reviewed weekly; meetings are problem-solving, not reporting
- •Weekly cadence reduces presentation theater and speeds course correction
- •Transparency culture: clear yes/no decisions, accept false positives for speed
- 40:18 – 50:18
US market, remote work, and hiring for growth: what matters most
Antoine discusses why the US is structurally different (credit-led) and why fintech penetration can be lower in mature economies. He then explains Revolut’s remote-work tradeoffs, and closes with hiring philosophy: prioritize specialists, assess problem-solving ability, and build analytics before spending.
- •US differs from Europe: credit-centric dynamics
- •Observation: mature economies can have lower fintech penetration (e.g., US, Germany)
- •Remote pros: meeting efficiency and better participation equity for introverts
- •Remote cons: reduced human contact; limited in-person time (annual meetup)
- •Hiring: use problem-solving case studies; decompose problems into granular parts
- •Common mistake: hiring ‘growth managers’/generalists instead of specialists
- •Advice: build analytical framework before spending money
- 50:18 – 1:01:48
What Revolut still must optimize: brand as product-led, plus quick-fire takeaways
Antoine says Revolut’s biggest remaining opportunity is brand—positioning and inspiration—and argues brand starts with product experience, with marketing as amplifier. In the quick-fire, he reiterates what persists (performance marketing), what’s harder now (virality), critiques privacy-driven platform narratives, and shares product ambitions like mortgages.
- •Biggest gap: maturing Revolut’s brand marketing and positioning
- •Brand is product-led early; marketing amplifies later
- •Early-stage founders often blame marketing when product isn’t truly resonant
- •Performance marketing remains key; virality is harder than it used to be
- •Critique: ‘privacy’ sometimes used to push unmeasurable brand spend
- •Future products: mortgages (and interest in student loans)
- •Admired strategy: Nubank’s depth in Brazil and expansion momentum
