Last summer I not only joined the financial services team at GOJEK to lead the Business Intelligence & Underwriting functions, I also relocated across the Pacific from California back to Jakarta, a place where although my entire family is based, I last lived over a decade and a half prior during my elementary school years. I wrote a post last year detailing why I moved and why I believe more members of the Indonesian diaspora should as well. In this post, I will detail some of the learnings from my last 52 weeks since taking the leap (NOTE: #hashtags represent GOJEK core values).
Even at a start-up (within a start-up) where things oftentimes feel like they’re moving at hundreds of kilometers per hour (yikes, I’ve been away from the US long enough to default to the metric system), pacing is extremely important. The spectre of burn-out is punctuated in the chaotic capital of a developing country, so don’t come out of the gates too fast and sprint a marathon… I learned this the hard way—quite literally—whilst running my very first marathon in Borobudur and pling both my calf muscles trying to hit an unrealistic target during the latter half of the race
Though this is a physical example, the same principle can be applied to working at a developing country start-up. Driven by my passion for financial inclusion and novelty of being in a new country, I over-committed to too many things in the beginning of my Jakarta 2.0 chapter: in addition to my core work at GOJEK, I was consulting a social entrepreneur working on waste banks for Ashoka, advising my brother how to scale his social enterprise, cross-fitting at 6am everyday, and training for a marathon.
You can imagine that this lethal, overly idealistic millennial cocktail burnt me out after around half a year. One thing that I’ve learned from running is that pacing is much easier in groups, which leads me to the second point…
2. Human capital is everything. Your team makes or breaks you, invest in them. #CollaborateWithCompassion #EarnYourTitle
When I first started, I was leading a team that consisted of me, an analyst and an intern. While small, I was fortunate enough to be working with colleagues whom I consider better than me at a few key functions. In a quarter, we somehow managed to set-up the entire reporting infrastructure to support the launch of our user credit product, underwrite a dozen batches of financial products to our drivers, and deep dive on merchant analytics to launch and scale our direct working capital loans. Today, I have the pleasure of leading a team of 10+ (and PS I’m hiring more).
That philosophy of hiring people better than yourself is even more cemented into our team’s DNA now. Each one of us has our own unique core competencies and we’ve cultivated a culture of learning from each other in a symbiotic way. One person might be a Kafka architecture expert, whilst another is a Python wizard, and another is a Tableau grandmaster—these diverse blend of skills are required to operate effectively as a holistic team. Additionally, I strongly encourage continuous learning. Particularly when you are working for a fast-paced tech-based start-up, continuous learning is not just a fluffy HR benefit, it’s a strategic necessity. This is why I’ve strongly encouraged all members of my team to improve their knowledge of statistics, Python, machine learning, etc. via MOOCs like Coursera or through trainings with our strategic partner/investor, Google. That being said, another way to learn brings me to my third point…
3. Failure is the best teacher. Create a safe space for mistakes. #CriticismIsAGift #ItsNotAboutYou
Creating a safe space for f*ckups and cultivating trust is key to high-performing teams. My learning whilst at GOJEK has been orders of magnitude higher than when I was in Silicon Valley; this fact surprises most folks who consider Silicon Valley to be the mecca of tech/the Athens of the 21st century/the El Dorado of start-ups. However, when you marry the DNA of Silicon Valley with values that embrace failing fast in an emerging market context, the pace of learning is subject to power laws.
Analytics, Business Intelligence & Data Engineering are complex things that aren’t easy, if they were there wouldn’t be an entire industry dedicated to [insert data function here]-as-a-service. Add a dollop of speed and scale, and you have a scenario that punctuates error rates. Whether data loss due to a backfill gone awry, an illogical scatterplot due to a LEFT JOIN instead of JOIN, or an experimental design muddied by a flawed control group, these are just a few mistakes that I’ve embraced as learning opportunities. In general, better structuring and planning can prevent a lot of data-related mishaps, however, this brings me to my fourth point…
4. Plan flexibly and overcommunicate. Be prepared to be unprepared. #AlwaysBePrepared #CommunicateWithPurpose
Leading a Business Intelligence division for an early-stage FinTech start-up (within a start-up) using perfect agile-scrum project management sprint cycles is often not possible and frankly not even desirable at times. Because Business Intelligence is neither a pure technical nor pure business function, deciding upon an appropriate project management framework is not always clear; I found myself asking: should we use a more strict Scrum framework commonplace in most software engineering teams? or should we use a looser Kanban framework commonplace in manufacturing?
Implementing Scrumban was a great decision, granting us the efficiency and focus of a Scrum, whilst allowing for the flexibility of a Kanban methodology. The key ingredient to really tying this together however is having a clear process, transparent channel of information, and tending towards over-communicating with all of our stakeholders. Though we’ve experimented with multiple methodologies, initially a proponent of the flexibility of Kanban, then an adherent to the rigidity of Scrum; at the end of the day, you need to constantly gather feedback on what works and what doesn’t work for your team specifically, in order to continuously improve. After closely monitoring the amount of points that my team was able to deliver against various frameworks, it became evident that Scrumban was the most efficient, hence my fifth (and final) point…
5. Data can be misrepresented, but it cannot lie. Inculcate a culture of healthy skepticism yet ultimate belief in data as an arbiter. #BeAScientist #StandUpForWhatYouBelieveIn
The best, most invigorating discussions that I’ve had in GOJEK (and frankly throughout my career) have been the ones where there is a strategic decision to be made—be it a product feature, a new product launch or a new marketing campaign—the stake of which rests on a careful, democratic examination of the underlying data wherein the decision-makers are informed, respective of the data, and yet feel empowered to debate on the merits of a decision. For instance, when deciding whether to offer a free trial as a way of boosting our quality MAUs (and ultimately, profitability) a sprint-long examination of various cross-tabs of adoption & retention data alongside the product team, led to decision to offer a single month free trial to bolster MoM retention. I cannot underline how key having this data-driven culture is. You’ve probably heard of this quote:
“What doesn’t get measured, doesn’t get managed”
— Peter Drucker, Professor of Management
But I recently heard this one which is even better:
“If we have data, let’s look at data. If we all have opinions, let’s go with mine.”
— James Barksdale, CEO of Netscape
The second quote suggests the primacy of having data as the ultimate arbiter, as being data-driven is much more than just having and seeing the data, it’s trusting and relying on data to make decisions. The latter is a subtle yet important distinction. One of the first actions I took when I joined the team was to host an org-wide session titled “Data-Driven Experimentation: applying dry statistics to not-so-dry situations” in order to demonstrate how to use mathematical techniques to drive decisions, optimize operations, and maximize marketing, to democratize data-driven decisioning (alliteration FTW), both my team and the broader GOJEK BI team have hosted trainings for non-BI folks through one-off SQL trainings and larger BI University courses. Such sessions are key for cultivating a data-driven culture and empowering all employees—including non-BI folks—to use more evidence in their decision-making.
If you’re reading this, thanks for making it to the end! The journey thus far has been illuminating, humbling, and transformative; if you’d also like to accompany me on this journey, GOJEK’s financial services and payments teams are looking to add more data-driven & mission-driven folks to its ranks, email me at max.hasan at go-jek.com with your CV and three ways you embody 3 of the 10 GOJEK values hash-tagged above.