4 guests 5 episodes 2,369 words

The Question

When is the right time to build a dedicated growth team?

When should a startup hire its first growth person or build a dedicated growth team? What signals indicate you are ready -- and what happens if you get the timing wrong?

The 4 Positions

1 / 4

Only After Product-Market Fit -- Growing Before PMF Is Pouring Water Into a Leaky Bucket

Sean Ellis

Sean Ellis, who coined the term "growth hacking" and developed the 40% "very disappointed" test now used by thousands of startups, has the most conservative and most widely cited position. He built his framework at LogMeIn, where the product initially tested below the PMF threshold. Rather than investing in growth, he repositioned around antivirus (what the "very disappointed" users actually valued), streamlined onboarding to deliver that value first, and within two weeks moved the score to 40%. Six months later it reached 60%.

"The question is, how would you feel if you could no longer use this product? Once you got a high enough percentage of users saying they'd be very disappointed, most of those products did pretty well. If you felt too low, those products tended to suffer."
Sean Ellis
▶ 00:00:07

Ellis's prescribed sequence is rigid: (1) find PMF, (2) understand the value you deliver by studying the "very disappointed" cohort, (3) build a north star metric that captures units of value delivered, (4) diagram a value delivery engine, and only then (5) systematically grow. Skip a step and you are growing something that does not retain.

"Step one would be understand it. Step two for me is then figure out a metric that essentially captures units of that value being delivered. And so when I think about a north star metric, that's what I'm thinking about is something that reflects how many people are coming in and experiencing that product-market fit experience."
Sean Ellis
▶ 00:49:45

His deeper point is about a failure mode that is nearly universal. Most startups reverse the sequence: they measure (analytics dashboards) before they understand (why users love the product), and they try to grow before they have done either. The growth hire who arrives before this foundation exists will inevitably optimize for surface metrics -- signups, page views -- rather than deep ones like retained users who get value.

"Just ignore the people who say they'd be somewhat disappointed. They're telling you it's a nice to have. If you start paying attention to what your somewhat disappointed users are telling you and then you start tweaking onboarding and product based on their feedback, maybe you're going to dilute it for your must have users."
Sean Ellis
▶ 00:00:25

When This Applies

No dedicated growth hire. Your product person should run the 40% test and instrument basic analytics. Growth investment is premature. This is Ellis's sequence, reinforced by the Verna/Wang newsletter post.

Hire Growth Early, But Scrappy -- Match the Hire to Your Stage

Crystal Widjaja

Crystal Widjaja, who built and led the growth team at Gojek during its explosive scale from 4,000 orders per day to Southeast Asia's largest super app (completing more rides per day than Lyft and more food deliveries than GrubHub, Uber Eats, and DoorDash combined), offers a pragmatic counter: you need growth thinking early, but the right kind of growth person matters more than the timing.

"I felt like it was a problem that was very solvable. And we ended up renting a stadium to just hire 60,000 drivers in a couple of weeks. So I think looking back, it was certainly a risk. When I got there it was in a house and I realized I've probably made a huge mistake, but we were growing very quickly already, even at that small scale of 4,000 orders per day."
Crystal Widjaja
▶ 00:00:00

The key insight: early growth hires need to be scrappy generalists who can instrument data, run experiments, and make decisions with incomplete information. You do not need a VP of Growth. You need a curious operator who can touch data, product, and marketing simultaneously. Hiring a senior growth leader before you have the data infrastructure for them to work with is like hiring a pilot before building the plane.

Widjaja also brings a critical nuance about what the first growth hire should actually understand about data. In her Reforge essay "Why Most Analytics Efforts Fail," she argues that most teams treat metrics as entertainment rather than actionable information:

"Do not treat metric gathering as entertainment. It's not there for you to be like, 'Oh, that's interesting, how novel,' and then not act on it. So real news is information that changes what you do in the real world."
Crystal Widjaja
▶ 00:40:18

When This Applies

Hire a data-oriented generalist early. The feedback cycle is fast enough that understanding and building can partially overlap, but you still need someone who instruments before they experiment.

The AI Era Changes the Timeline -- Feedback Cycles Have Collapsed

Elena Verna

Elena Verna's argument is the most forward-looking and the most destabilizing to conventional wisdom. As Head of Growth at Lovable, she is operating at a company that reached $200M ARR before its first birthday with fewer than 100 people. Her claim is not theoretical -- it comes from watching the old rules break in real time.

"It's the feedback cycle. It's just completely collapsed. You can go from idea to some product that is functioning to user feedback within a day if you want to."
Elena Verna
▶ 00:44:18

If you can build a functioning product in a day and get real user feedback immediately, the argument for waiting until PMF weakens. You can run growth experiments as part of finding PMF, not just after.

Verna goes further, admitting that most of what she learned in nearly two decades of growth work simply does not apply:

"I feel like only 30 to 40% of what I've learned in the last 15 to 20 years of being in growth transfers here because we just need to invest in such bigger bets, and innovate, and create new growth loops here."
Elena Verna
▶ 00:00:14

She also reframes the quality bar itself. In the AI era, "viable" is table stakes. The bar is lovability:

"I call it minimum lovable product. It shouldn't be minimal viable product anymore. Viability is left back in 2010s. Now it's minimal lovable product."
Elena Verna
▶ 00:43:46

When This Applies

First growth hire: scrappy generalist in the Crystal Widjaja mold who can instrument, experiment, and analyze. Not a VP of Growth. Start with Bangaly Kaba's understand work.

Start with Understanding, Not Building

Bangaly Kaba

Bangaly Kaba -- early growth PM at Facebook, Head of Growth at Instagram (scaled to 1B+ users), VP of Product at Instacart, now Director of Product Management at YouTube -- brings the most operationally rigorous framework. His core concept is a deliberate inversion of what he calls the anti-pattern:

"Someone says, 'Hey, you know what? This would be great to build.' And you go pull data to go justify why that would be great to build. Call that identify, justify, execute. First you have to really understand from first principles what is actually going on. So understand, identify, execute."
Bangaly Kaba
▶ 00:00:29

The "understand" phase is not background research that happens informally. It is named, assigned work on the team's roadmap. At Instagram, understand work meant data science collaborations to map the social graph, instrument logging, and analyze what predicted long-term retention. At YouTube, when Kaba arrived, teams allocated 60% to execution and 40% to understand work. As understanding deepened, the ratio shifted:

"When your team does not fully understand a problem space, the balance of work tends to be higher. I think when I started at YouTube we were doing 60% execution, 40% understand work. And over time, as we understand more and more, the mix shifts. Now my teams are probably doing 80% execution, 20 to 15% understand work."
Bangaly Kaba
▶ 00:31:47

The payoff was extraordinary: at Instagram, 15 growth teams ran 12-20 experiments per quarter per team with a 60-70% positive and shippable rate:

"We had 15 teams and we might've been running 12 to 20 experiments a quarter a team, and I would say probably 60 to 70% of them were positive and shippable, which is incredible."
Bangaly Kaba
▶ 00:30:29

The practical implication for hiring: your first growth person should be someone comfortable spending three months doing nothing but analysis before building a single experiment. Most growth candidates want to "show impact" in their first quarter. The right first growth hire wants to understand the system before touching it.

When This Applies

Start with either a dedicated growth PM or a cross-functional tiger team, per Hila Qu's framework. The growth hire should combine product and marketing skills, and data infrastructure is the prerequisite, not the afterthought.

Evidence from the Archive

GrowthHackers

Dropbox: Ellis helped build growth on top of confirmed PMF, not before it

LogMeIn tested below 40%, repositioned on antivirus, streamlined onboarding, hit 40% in two weeks and 60% in six months -- eventually a $4B+ exit

Sean Ellis · Author of Hacking Growth, coined the term 'growth hacking' ▶ 00:00:07

Lovable, formerly Miro, Amplitude, Malwarebytes, SurveyMonkey

Lovable hit $200M ARR in under a year with fewer than 100 people, with growth still accelerating

A Lovable team member vibe-coded a full feature in two weeks that previously would have taken months of research, design sprints, and roadmap prioritization

Elena Verna · Head of Growth at Lovable ▶ 00:44:18

YouTube, Instagram, Facebook, Instacart

YouTube: started at 60% execution/40% understand, shifted to 85%/15% as understanding deepened

Instagram growth teams: 15 teams running 12-20 experiments/quarter with 60-70% positive/shippable rate, driven by understand work

Bangaly Kaba · Director of Product Management at YouTube ▶ 00:00:29

Gojek, Kumu

Gojek scaled from 4,000 orders/day to surpassing Lyft in rides and DoorDash in food delivery

Gojek: rented a stadium to recruit 60,000 drivers in weeks -- a growth tactic no playbook would suggest, requiring zero-to-one creativity

Crystal Widjaja · CPO at Kumu, former growth team leader at Gojek ▶ Watch

The Synthesis

The four positions form a sequence that most growth advice collapses into a single recommendation. The real framework is temporal and conditional.

Three Distinct Activities

Why does most growth advice collapse a sequence into a single recommendation?

Infrastructure Before Team

What must exist before a growth hire can succeed?

Hire Type Matters

Why do growth hires from big companies often fail at startups?

'Growth' is not one activity -- it is three distinct activities (understand, build, optimize) that become appropriate at different stages. You can start understanding before PMF. You should not start optimizing until well after it. Building growth features requires enough understanding to know what to build.

Product-led growth is data-led growth. Without data infrastructure, experimentation platforms, and lifecycle marketing tools, a growth hire has nothing to experiment with. The infrastructure must precede the team. A growth leader will take at least 6-12 months to understand the local problem fully.

Early growth hires from large-company growth teams often struggle at startups because they are optimizers, not builders. They know how to run experiments on existing traffic but not how to create traffic from zero. Early growth demands zero-to-one creativity, not optimization playbooks.

Which Approach Fits You?

Answer 3 questions about your situation. We'll match you to the right approach.

Question 1

Have you achieved product-market fit?

Question 2

How fast is your feedback cycle?

Question 3

What does your data infrastructure look like?

Notable Absences

The Bottom Line

There is a second hidden lesson across these conversations: the *type* of growth hire matters as much as the timing. Early growth hires who come from large-company growth teams (Meta, Google, Uber) often struggle at startups because they are optimizers, not builders. They know how to run experiments on existing traffic but do not know how to create traffic from zero. Crystal Widjaja's "rent a stadium" story is the proof case. That is not a tactic from any growth playbook. It is the zero-to-one creativity that early growth demands.

Elena Verna is right that AI compresses the timeline. But compression does not eliminate the sequence. You still need to understand before you build and build before you optimize. It just all happens faster.

Sources

Sean Ellis — "The original growth hacker reveals his secrets" — September 5, 2024
Crystal Widjaja — "How to scrappily hire for, measure, and unlock growth" — July 31, 2022
Elena Verna — "The new AI growth playbook for 2026" — December 18, 2025
Bangaly Kaba — "Unorthodox frameworks for growing your product, career, and impact" — May 26, 2024
Elena Verna & Andrea Wang — "How to hire your first growth team" — August 22, 2023
Hila Qu — "Five steps to starting your product-led growth motion, part 2" — January 17, 2023