Ed Baker on Why Growth Leaders Should Do Less, Not More

The Counterintuitive Math of Fewer Bets

Most growth leaders approach their role like venture capitalists: place a hundred bets, hope a few hit. Ed Baker learned to do the opposite. During his tenure as VP of Product and Growth at Uber from 2013 to 2017, he discovered something that contradicted conventional wisdom about experimentation velocity.

"When I look back at my time at Uber, every year we maybe had 1 or 2 things like that that accounted for more growth than everything else combined," he explains. "All the hundreds of other A/B tests that we were doing, put it all together and that didn't drive as much growth as one little thing."

One of those "little things" was a button in the rider app that said "want to make some money, become a driver." The addition took minimal engineering time. The impact was seismic: "Suddenly we had as many new drivers signing up from that button as we did from all of our paid marketing efforts combined." This wasn't luck. It was the result of ruthless prioritization around what Baker calls "needle movers"—interventions that create step-change growth rather than incremental lifts.

The insight scales backward too. At Facebook, where Baker ran international growth after the company acquired his startup Friend.ly (which had grown to 25 million users without marketing spend), a similar pattern emerged. In Japan, Facebook was losing to local competitor Mixi. The team noticed Japanese users weren't inviting friends—a critical step in the viral loop. The cultural diagnosis: invitations felt rude. The fix: rename "invites" to "announcements." Conversion rates spiked. Japan went from "one of our slowest growing countries to one of our fastest growing countries," and Facebook surpassed Mixi shortly after.

The North Star Metric as Organizing Principle

Baker's framework for identifying needle movers begins with a single question: what is the one metric that encapsulates the health and trajectory of the entire business? At Facebook, the answer was monthly active users (MAU). At Uber, it was weekly trips—chosen because "every trip requires a rider and a driver," making it a true two-sided marketplace metric rather than a vanity number.

If you're not working on something that's moving this metric in a positive way, you're probably not working on the most important thing.

— Ed Baker

This North Star becomes the forcing function for prioritization. Baker's approach: make assumptions about conversion impact, map that to North Star movement, then stack rank. "Even if it's a nice to have, it's not going to move the needle as much. So let's not spend resources on that thing." The discipline sounds simple. Execution requires saying no to dozens of projects that would feel productive but don't move the number.

The metric also structures team cadence. At Facebook, the growth team met weekly to answer one question: "How much has MAU grown or not in the past week and why?" That rhythm—weekly review, relentless focus on a single number—creates accountability and surfaces anomalies fast. When Japan's invite conversion dropped, the data flagged it immediately.

Five Functions, Not Five Hundred Experiments

When Baker started Uber's growth team in 2013, he had five people. His mental model, imported from Facebook, was to staff five critical functions: product, design, engineering, marketing, and data science. "Mike, you're going to run product, Curtis, you're going to run engineering, and so on," he recalls. Some had never done their assigned function before—Uber's first growth marketer "had never done marketing before"—but the structure created clarity.

This cross-functional team reported directly to the CEO, not to product or marketing. Baker views growth as "a cross-functional team effort to move the metric that ultimately drives the growth of the business." That includes operations. At Uber, onboarding drivers required in-person vehicle inspections and background checks. Growth wasn't just pixels and targeting parameters; it was the entire funnel, including the parts that didn't scale.

I think of growth as just like whatever you have to do to grow that number faster falls under growth, and that can include marketing, that can include product. In fact, a lot of times it is product-driven changes.

— Ed Baker

The tension Baker never fully resolved: overlap between growth and product teams. As both scaled to hundreds of people, friction emerged. "Anytime you have areas of overlap in any business, that can create some friction. And I think that's also when politics and that kind of stuff appears." His conclusion isn't prescriptive. He's seen companies succeed with growth inside product, and he's seen companies—like Booking.com—operate with only a growth team. The structure matters less than the leader and the clarity of the North Star.

Retention Over Acquisition at Scale

Baker's experience at Facebook upended an assumption he carried from his early viral loops. "When I got there, we focused so much more on retention and engagement than acquisition," he says. "As you get bigger, those two metrics of retention and engagement actually become a lot more important to maintaining growth than new user acquisition."

The logic is simple once you see it: if you have 100 million users and 10% churn monthly, you lose 10 million users. Acquisition has to replace that 10 million just to stand still. Reduce churn by 1 percentage point, and you've added a million users without spending a dollar on ads. The marginal return on retention compounds. The marginal return on acquisition usually degrades as you exhaust cheaper channels.

At Uber, this translated into a supply-side obsession. "We needed to add more drivers in order to grow more so than riders," Baker explains. The bottleneck determined the focus. In a two-sided marketplace, growth isn't symmetric. You optimize the constrained side, because the other side will follow. That's why the "become a driver" button in the rider app was a needle mover: it relieved the binding constraint.

When to Build the Growth Team (And When to Wait)

Baker is blunt about timing: hire for growth only after you have product-market fit. "I think one mistake startups sometimes make, and I've made this mistake myself as well, trying to focus on growth too early." He draws a hard line: product-market fit is building something people love. Growth is adding fuel to that fire. Conflating them leads to scaling a product nobody wants.

An analogy I like to use is like drilling for oil. No matter how good your drill is, if you're drilling in the wrong place, you'll never hit it. You don't want to start growing if you're drilling in the wrong place.

— Ed Baker

His litmus test is qualitative, not quantitative. At Uber, he remembers his first trip as "a magical experience." Drivers told him they were making more money than ever before in a flexible way. "To me, that's the product-market fit." Only then did it make sense to build a growth team.

The timing question extends to structure. Baker advises founders to first decide where growth lives—inside product, alongside it, or as the only team. At Whoop, a company he advises, growth lives within product. The choice depends on the leaders you have and the org's muscle memory. What doesn't work: hiring a "growth leader" to find product-market fit for you. That's the founder's job. Growth scales what's already working.

The Viral Loop That Started It All

Before Facebook, before Uber, before the frameworks and the North Star metrics, there was datesite.com. Baker built it in 1999 at Harvard—a crush-matching site where you entered email addresses of people you liked, and if they liked you back, you'd both get notified. He launched it on Valentine's Day. Within one week, a quarter of Harvard had signed up.

"For me, that was my first time building a website, but also seeing exponential growth on the internet," he recalls. He was studying physics at the time. The experience sent him down a different path: measuring viral loops, calculating K-factors, understanding the conditions under which K exceeds 1 and growth becomes exponential. Facebook's app platform became his laboratory. He built "several silly apps that all grew to millions of users just through this viral loop."

We sold the company to Facebook. I ended up running the international growth team there. And at Facebook, I really learned what it's like to work on a larger growth team.

— Ed Baker

Friend.ly, the social Q&A site he co-founded after Stanford Business School, hit 25 million registered users with zero marketing spend. The pattern held. Growth, for Baker, has always been about understanding the mechanics of virality, identifying the one or two levers that matter, and ignoring the noise. Twenty years after datesite.com, the philosophy hasn't changed. Just the scale.