Antoine Le Nel on Why Revolut Killed CAC Targets and Runs No A/B Tests

Why Revolut Banned CAC Discussions

Most growth teams obsess over customer acquisition cost. Antoine Le Nel thinks that's exactly why they fail.

When Le Nel joined Revolut as Chief Growth and Marketing Officer, he brought a playbook from 7 years at King (makers of Candy Crush) where CAC was never mentioned. The reason? Optimizing for CAC means you always chase the cheapest, worst cohorts.

There's no CAC discussion at Revolut. Why? We only talk about ROI. Because if you talk about CAC, you'll always acquire the worst cohorts.

— Antoine Le Nel

Le Nel's first move was to increase CAC significantly while improving ROI. The team stopped asking "how cheap can we get users?" and started asking "how valuable are the users we're getting?" TikTok might deliver lower CAC than Instagram, but if the cohorts don't convert to revenue, the metric is meaningless.

The shift requires better analytical models. Le Nel's team built cohort prediction frameworks that identify high-LTV users early, sometimes within days of signup. Those models work across geographies because human behavior around financial products is more similar than different.

When A/B Testing Is a Competitive Disadvantage

Le Nel arrived at Revolut from King, where the company ran 1,000 parallel A/B tests in optimization mode. At Revolut, he found almost no testing. His first reaction was shock. Then he understood why it was brilliant.

There are three reasons Revolut skips A/B tests. First, good product managers should know what the right product is without needing data validation. Second, when you're building for 10x improvements, you don't need a test to see the difference. Third, A/B testing destroys time to market.

If you're a good product manager, you don't need an A/B test. You should be able to know what is the right product. When you do a 10x, you don't need an A/B test to measure it. You do it, you see it goes up, it's good.

— Antoine Le Nel

The distinction comes down to lifecycle stage. King was mature, optimizing incremental lifts on a proven game. Revolut is still in creation mode, where speed matters more than precision. If something doesn't work, they cut it in 2 weeks. A TV campaign scheduled for 3 months gets killed after 14 days if the data is bad.

The company maintains velocity through radical decentralization. Each team operates as a self-sufficient startup with minimal dependencies. This creates what some would call silos, but Le Nel calls it the only way to stay fast at scale.

The Attribution Frameworks No One Else Has Cracked

Growth is a bidding war. The winner isn't who bids highest, but who can measure ROI accurately enough to know how high they can bid. Le Nel's competitive advantage at both King and Revolut has been cracking measurement frameworks before competitors.

At King, they figured out TV attribution years before other mobile gaming companies. That meant King could scale TV spend while competitors stuck to digital channels with clearer attribution. The same pattern repeated with iOS performance marketing post-ATT. Companies that solved iOS measurement kept spending; everyone else shifted budgets to Android at 2-3x the rate.

Revolut's attribution plays are more creative. They advertise on airport jet bridges, then place vending machines nearby where travelers can grab a Revolut card instantly. Scan the QR code, create an account, and the anonymous card links automatically. Perfect attribution in a historically unmeasurable channel.

Growth is a bidding war. Your competitive advantage will always come from your ability to measure more than your competitors, because then suddenly you're going to do that thing.

— Antoine Le Nel

Le Nel structures the full funnel into three layers. Upper funnel is brand marketing with high reach but loose attribution. Lower funnel is performance marketing, affiliates, and referrals with full attribution. Mid-funnel is where Revolut wins: activities like influencer partnerships that drive immediate revenue while building brand consideration. High reach, trackable revenue, deeper content than 6-second ads.

Why Localization Is Overrated and Beliefs Are Dangerous

Romania and Norway became Revolut's number one markets using the identical strategy. Same creatives, same approach, same everything. Le Nel's experience at King proved that localization beyond language is mostly wasted effort.

The exception? Japan. Every other market responds to the same creative, same messaging, same product positioning. The extra performance from hyper-localization doesn't justify the complexity cost. Consumer interests are global, not local.

This philosophy extends to Le Nel's broader growth approach: he has no beliefs. Not in brand marketing, not in influencers, not in any specific channel. Formula One sponsorships, celebrity partnerships, billboards—none of it matters unless the data proves ROI. When campaigns with big-name celebrities flopped because they weren't social-savvy, Revolut killed them fast.

I have no belief in anything. I don't necessarily believe more in brand than in influencers than in this and so on. Whatever works works for me.

— Antoine Le Nel

The growth engine breaks down when belief enters the equation. Operators bring playbooks from previous companies and assume what worked before will work again. Le Nel brought his King playbook to Revolut but changed it substantially. He'd never been a fan of upper-funnel brand marketing at King. At Revolut, he tested it, found it worked, and scaled it. No loyalty to past patterns, only to present data.

The Snack Strategy That Beats Switching Costs

Traditional banks force a binary choice: switch or don't switch. Revolut treats itself as a snack. Try it for holiday spending. Maybe buy some crypto. Perhaps open a savings account with better rates. Three years later, users realize they've organically migrated their primary banking without making a conscious switching decision.

This changes everything about how Revolut measures success. The North Star isn't "percent of users who are primary account holders by quarter." It's the progression rate: how many users are moving along the journey from single-use-case to multi-product to primary bank, at their own pace.

Le Nel tracks three metrics: volume of new users, ROI at acquisition, and progression toward primary usage. Payback periods stay stable across markets because the models are standardized. The growth engine is steady, accelerating week over week, with no plateaus visible even in mature markets.

The hardest part of the role isn't building the engine. It's evolving it continuously while it's working. Le Nel compares it to the cow getting eaten from the back—by the time you notice something's broken, you're already gone. Targets at Revolut are set ambitiously high, rarely hit at 100%. The team operates in founder mode on steroids, driven by founders who maintain the same intensity a decade after launch. The question isn't whether they'll slow down. It's how long they can sustain the acceleration.