It’s been a few months since I started working at an early/growth stage SaaS startup as their first data hire. My title is Staff Data Engineer, but I’m also a data analyst, data wrangler, data strategist, etc. I’m working on various data projects across departments in the company. I can probably say I’m a De Facto head of data, given what I actually do day-to-day.
In this post, I’d like to reflect on what’s the ride been like and what I learned so far.
Here’s the sections in this post:
Balancing Workload
First of all, this is obvious, but there is a lot to be done at a startup. In my particular case, When I first got hired, there was minimal analytics infrastructure, governance, strategy, and immatrue processes that generate messy data (perhaps a typical startup setup). And it sounded like they were not even looking for a data person at the time I was hired, I can see I just popped up within their network and they figured I’d be a good fit.
Ever since I was hired, I worked/am working on things like:
- Building the company data strategy
- Coming up with priorities and roadmap for internal data analytics
- Building the entire analytics infrastructure from scratch
- Building solutions for ad-hoc analytics
- Building automated-dashboards for departments
- Helping with data room/data portion of investor deck
- Developing metrics definitions
- Implementing security data practices (e.g. RLS, OLS, data masking)
- Building documentations on data/analytics assets
- Helping with source data cleanups
- etc
I’ve been learning a ton and working hard. Not that I’m required to work over 40hrs/week though (in fact the CTO is GREAT. He understands the IC‘s point of view, because he’s been an IC for a decade or two, and he doesn’t think we should be working more than 40hrs/week to get things done, is at least what I understand from what he says). That said, I work on work stuff outside of 9-5 from time to time. It’s more like it’s because if I get more things done, I’d have more, direct impact on the business. My work is directly related to improving the business, which motivates me a lot in working for this kind of startup, a small, yet growing company.
Also, one great thing about being the first data hire is you have flexibility in choosing the data tools. Because I have the “best” idea on how the data work needs to be done, the only thing I’d need to do is to convince my boss, finance or the exec team (I guess that’s the hard part). And when a tool is something you can self-host? You may not even need permission from the management. Having this flexibility or autonomy actually makes my work easier and more enjoyable.
In contrast, here are some things that I feel have been challenging:
- Visibility
- Not everything you do are visible. I guess this isn’t specific to working at a startup, but when you build a dashboard/analysis, you’d have to build other things like analytical models and other infra pieces, which are not necessarily “visible” to the business. Also if I build a solution for Customer Success for example, it’s not that other exec members automatically know that part of my work.
- Lack of documentation
- My dream is to have every single column/table documented in prod database (if it’s an in-house-app), but it’s not the case mostly. And it’s not trivial to figure out what each column means exactly and how we can use it for analytics.
- Everything moves faster (e.g. source data structure)
- New columns added every day, column meanings change, a new table gets added that would change the structure of our analytical models, things like these happen frequently. That’s fine by themselves, but at a startup you don’t have weeks or even days to think on how these changes should be implemented in the current the structure of your analytical models that support reporting.
- Clear boundaries and areas of responsibility
- I think it’s my title (data engineer) that makes people confused. And this is also related to visibility. The problem is that people don’t always have a clear picture of my areas of responsibility, or what I do.
To summarize, let me say that building everything on my own has been super fun. Though it’s not easy balancing high-level items, infra work, and building user-facing solutions. Also, a lot of decision need to be made during the day, one of these cases being deciding to build something in a scalable manner or just get it done quick. I’m beginning to realize decision fatigue might be a real thing.
Data Strategy, Roadmap, Hiring Plans
When you’re working as a data person, you’d typically think things only inside the box. But when you’re a solo data hire at a company, you’re the manager, lead, and even de facto head of data, even though your title doesn’t say it explicitly. So, you not only do hands-on data work, but you also work on the overall data strategy, roadmap or project priorities, and hiring plans. And sure enough, all that comes with a ton of communications across teams and departments I need to manage.
When you’re buried by day-to-day projects, it’s easy to forget the high-level views that are crucial you want to be aligned on for everything you do. Because that’s what determines the level of impact your project would have on the business. It’s not been easy to figure out when to work on these high-level items. As time of this writing, about ~9 months-in into my role, I got a data strategy in place, and roadmap / priorities for 2026 is being currently worked on. I have a hiring plan but not too official.
One thing I’ll do differently if I did work as a solo data person again at a startup is that I would start working on the data strategy and roadmap pieces early. I waited until I got a good grasp of what data looks like and business needs are, but I think if I had started working on them a bit earlier, it’d prioritize my projects better and give me exec-visibility early in my journey.
From data strategy to roadmap and hiring plans, I initiated the development of everything and nothing came from the top. On one hand I kinda wished the exec team had a vision or rough plan on what the data team should do, because then I’d have an easier time building the strategy and plans. On the other hand, I liked how I needed to initiate all these pieces because that helped me own them, not just contributing.
Context switching is real though. Something like switching from hands-on work to working on priorities and writing documentation happens every single day. Now that I’m writing this, I think I should come up with a framework for context switching that would reduce the friction/burden on my brain.
New Perspectives
Company culture matters a lot. Luckily the startup I’m at has a great culture, value and vision for its employees. The management has never been toxic or forceful in any way so far. And this is for sure reducing my stress on my delivery and daily communications. I’ll definitely look for the similar culture if I were to work for another startup in the future.
One of my goals in this role is to be a strategic partner to each department, not just a data person who passively takes orders. And while I feel I’ve done a good job on this goal overall, I think I could’ve done it better initially and I hope I’d be more “aggressive” and “vocal” in making the business better with data and analytics.
One thing I’ve noticed about myself is that I tend to shy away from getting attention in general. And speaking up in a meeting takes a good deal courage with some anxiety, especially if I’m trying to give an idea or opinion on how something should be done on the business side (since it’s not my territory). For that reason, oftentimes I rather want to just work on hands-on pieces of projects all-day and not discussing high-level items or coordinating cross-functional priorities in meetings. But I know the importance of getting myself out where people can see me and my work, and I don’t want to be some data engineer who just sits at his desk all-day and nobody knows what he’s doing. So, being kinda “shy” is something I want to change (and after all I may not appear shy to those I talk to). I still enjoy cross-functional communications though. And whenever I get talking with one of the business functions, I learn a ton about it and what’s important to that specific function, which has been an invaluable experience.
Summary
All in all, I really like being the first data hire at this startup. It helps me grow in ways I didn’t expect and I like how I can make an impact on the business in the way a typical data person wouldn’t be able to. I can see I stay long in this role, but you never know how things change at a startup.
P.S. I’m thinking to start taking consulting gigs on the side again soon. If you’re running a startup or know anyone who would benefit from meeting me, feel free to hit me up!