Growth-as-Cleanup-Crew Team Design
Point your first growth hire at known gaps, don't ask them to discover strategy from scratch.
- Difficulty
- Moderate
- Time to result
- ~months to results
- Steps
- 4
- Confidence
- 90%
A model for how to stand up and staff an early growth team. Growth exists to fill the acquisition, adoption, and engagement gaps the core product team leaves behind — it's a 'cleanup crew' that connects dots, not a group asked to invent strategy from zero. Because the physics of the business are largely fixed, the first growth hire should be pointed at the biggest known, workable gaps, and should be a statistically-literate person who optimizes for impact over flashiness.
Origin
Based on how growth was built at Gojek, which Widjaja calls a best-practice example; growth PMs were embedded into teams and later owned specific parts of the product stack.
Core principles
- 01Growth fills the gaps core product leaves behind — acquisition, adoption, engagement plumbing
- 02Don't set the bar impossibly high by asking a new hire to model everything from scratch
- 03Point growth at known, workable gaps where data already shows where the biggest problems are
- 04Hire statistically-literate people — sampling, selection bias, and correct measurement are the core skill
- 05Growth is a numbers game at scale: everything does something, so find the highest-impact thing now
- 06Growth is an opportunity-cost tradeoff against product time — bias toward quick, hacky wins
How to run it
- 1
Identify the gaps, then hire against them
Before hiring, map where the product visibly needs help — OTP/SMS delivery rates, first-time-user experience, onboarding education — and recruit someone to tackle those known gaps rather than to figure out what to do.
Pro tip Frame the role as 'here's where we need growth help, come tackle it,' not 'come figure out how to grow us.'
Watch out Asking a new hire to model and discover everything sets the bar too high; the physics rarely move, so hand them workable, data-backed gaps.
- 2
Hire for statistical literacy first
Prioritize people who can run the numbers — understand sampling, selection bias, and that a ratio on a tiny base can mislead — over the most experienced or flashiest candidate. Junior or senior both work if they focus on the right opportunity.
Pro tip Widjaja would hire someone willing to take an intro-to-statistics course over someone chasing the flashiest feature.
Watch out The worst hire is a growth person who thinks they're right, measures wrong, and optimizes the wrong areas.
- 3
Test experiment design, not live cleverness
In hiring, give a take-home case study asking how they'd know a claim is true and to design an experiment; look for random sampling and a measured, deliberate approach rather than 'build it and it'll obviously work.'
Pro tip Give ~5 days and set expectations (~4 hours of work) so you see their best quality; value candidates who admit they googled it and read white papers. Live formats don't reveal this well.
Watch out Beware candidates who bolt on many new tools that take six months to integrate — growth is an opportunity cost, so favor quick hacky wins.
- 4
Embed growth, or silo it only under strong PMF
Embed growth PMs into cross-functional teams so they own parts of the product stack. Keep growth as a separate side team only when product-market fit is so strong the org is scrambling purely on core features.
In the wild
Because Gojek grew so fast, the core team was still building basics (like phone-number masking), so growth took the gaps: choosing the right SMS provider per telco to land OTPs, and flagging drivers when a rider was brand-new so first-time users weren't stranded at pickup. These weren't core features but were adoption/engagement gaps.
→ Growth PMs synthesized the role by owning acquisition, adoption, and engagement plumbing, eventually owning specific parts of the product stack.
The team's first real growth experiment (while still the data team) connected a quick Python script to the Twilio API and SMSed drivers from an uploaded CSV telling low-acceptance drivers to accept their rides.
→ Acceptance rates rose ~2% across the board and more for brand-new drivers, feeding an improved driver-onboarding experience — a quick, hacky, high-ROI win.
Common mistakes
Expecting the first growth hire to invent strategy
Asking a new hire to model and discover the whole growth strategy sets an impossible bar; since the physics are fixed, they should be pointed at known workable gaps where data already shows the biggest problems.
Onboarding many new tools before shipping anything
A growth team can go terribly wrong buying tools that take six months to integrate and produce nothing meanwhile. Growth is an opportunity cost against product time, so quick hacky wins beat heavy tooling.
Is it for you?
Best for
Founders making their first growth hire or standing up an early growth function amid fast, messy scaling.
Not ideal for
Companies with such strong product-market fit that a siloed growth 'cleanup' team is sufficient and embedding isn't needed yet.
From the transcript
“we really called ourselves like we're the cleanup crew we pick up the pieces that were left behind we connect the dots”
“it's just setting the bar like too high to expect someone to come in and model everything”
“you have to have someone who knows how to run the numbers right”
“i actually look for that first principle bias so i'll give people case studies of like here's what we see how do you know that…”
“growth is an opportunity cost of time trade off with what you could have been doing to the product”
“the worst possible thing is to have a growth person who thinks they are doing the right thing and is measuring things wrong”
From the episode
How to scrappily hire for, measure, and unlock growth
Crystal Widjaja, Gojek and Kumu