Do the Job Badly to Learn What Excellence Looks Like
Founders in talent-poor markets should personally hold each role until they can recognise someone better
- Difficulty
- Moderate
- Time to result
- ~months to results
- Steps
- 4
- Confidence
- 90%
Aluwi held nearly every function at Gojek — co-CEO, de facto CPO, CFO, performance marketer writing push copy and uploading Facebook ads, even driver. The purpose was not scrappiness for its own sake: without a mature talent market you cannot hire by pattern-matching a résumé, so the founder must do the job, be bad at it, and thereby learn what good actually looks like — then hire someone orders of magnitude better.
Origin
Kevin Aluwi's practice at Gojek, born of Indonesia's thin talent market circa 2011-2015 where 'hire person X from company Y with job description Z' was not an available move. He distinguishes this from the driver role, which he did for empathy rather than benchmarking.
Core principles
- 01In a mature talent market you can hire on credentials; in an immature one you must hire on judgement, and judgement requires doing the job.
- 02Being bad at a job teaches you its shape faster than reading about it.
- 03The goal is a hiring benchmark, not permanent ownership — you're looking for someone orders of magnitude better than you.
- 04Some roles you take on for empathy, not benchmarking — be honest about which is which.
- 05First-hand experience of a role converts into product decisions no amount of research surfaces.
- 06This is founder work. It is largely not transferable to executives.
How to run it
- 1
Identify roles you cannot hire by pattern
List the functions where you can't credibly evaluate a candidate because you don't know what excellence in that role looks like in your market.
- 2
Do the job yourself, visibly and badly
Actually perform the work — write the ad copy, upload the campaigns, ship the push notifications. Aluwi calls himself 'an amateur performance marketer' doing exactly this.
Pro tip Seeing how bad you are at it is the mechanism, not a side effect — the gap between your output and what the job needs IS the definition of good.
- 3
Convert the experience into a hiring bar, then hire past yourself
Write down what the job actually entails and what a great outcome looks like, then recruit someone who can beat your version by an order of magnitude.
Watch out Holding the role past this point is ego, not strategy. The output of the exercise is a hire.
- 4
Separately, work the supply-side role for empathy
For roles your users or partners perform (driver, seller, support caller), do the job to feel the friction — not to benchmark a hire. Feed what you learn into product principles.
Pro tip The insight rarely lands instantly. Aluwi's driving experience only surfaced later, when driver teams pushed for waiting fees and multi-stop.
In the wild
In Gojek's early days Aluwi wrote push-notification copy, wrote ad copy, and uploaded ads to Facebook and Google himself, trying to optimise online marketing spend. He was not trying to be scrappy — he was trying to learn what the job was so he could evaluate a real marketer.
→ He built an internal benchmark for what good looked like in each function and used it to hire people who could do the work orders of magnitude better than he could.
Aluwi was the first driver on Gojek's car ride-hailing service. On one trip a customer came out with a giant bag, which he had to load into his trunk; mid-route she asked to detour to drop off her laundry, so he lugged the bag out and helped her, then continued to the mall. He earned very little for the trip.
→ Later, when Gojek's driver teams pushed for waiting fees and multi-stop support so extra work was compensated, Aluwi backed them — the driver app's product principles were shaped by his own unpaid detour.
Common mistakes
Hiring by résumé in a market with no résumés
The 'person X from company Y with job description Z' heuristic depends on a deep, legible talent pool. Where availability of experienced talent is low, it produces confident bad hires.
Confusing empathy tours with benchmarking
Aluwi is explicit that driving was not about learning what an excellent driver looks like — it was about empathy for the job so the product served it. Blurring the two produces founders who think a day of support tickets qualifies them to hire a VP.
Executives copying the founder playbook
Aluwi flags this as most relevant for founders and less so for executives — an exec doing every job is usually failing to delegate, not building a hiring bar.
Is it for you?
Best for
Founders building in markets with thin, illegible talent pools, who need to make senior hires in functions they've never performed
Not ideal for
Companies in mature ecosystems with deep talent markets and reliable hiring signal, or founders whose time is the binding constraint on the company's survival
From the transcript
“I do think it's really important to understand the work um that needs to be done in order to see what Excellence looks like”
“and seeing how bad I am at it uh allowed me to understand what good looked like”
“I held a lot of those roles just because you know I wanted to understand um every part of the business as best as I…”
From the episode
Taxi mafias, cash vaults, and 100% MoM growth: The story behind Southeast Asia’s biggest startup
Kevin Aluwi (Gojek)