Startup VS Corporate: Where to Work as a Data Scientist | by Conor O’Sullivan | Sep, 2022
The differences in culture, workload, mentorship, learning rate and access to data and tools
Walking into the office looked more like a day at the beach. Shorts and flip-flops. Was this startup culture or just Cape Town? From the CEO to the new hire (me), the entire 8-person team worked from the same room.
During my time there, I briefly consulted for one of South Africa’s telecommunications companies. The contrast was stark. I spent time at the head office in Johannesburg. 100 people in one department and closed shoes — how can I work under these conditions!? The CEO didn’t even know I existed.
Startups and large corporates are on the two extremes of working life. They can offer vastly different work and environments. These differences will ultimately drive your choice of where to work.
I want to share the differences I’ve experienced with a focus on…
- culture,
- workload,
- mentorship,
- learning rate,
- access to data and
- tools used to work with that data.
To end, I’ll touch on how this experience has shaped my aspirations as a data scientist.
At the startup, the environment was relaxed. No dress code (obviously) and the hours were flexible up until a point. If I wanted to arrive late I just worked later. No questions asked. These relaxed rules, extended to the way we interacted. We could talk freely and there was no obvious hierarchy.
For the majority of my corporate experience, I worked for a large bank in Ireland. That environment was less relaxed. Working hours were set and there was a clear hierarchy. Your manager needed to know if you’d be late for work.
There was a dress code but it wasn’t too strict. Jeans and hoodies for the most part. Occasionally, we spent the day at the grown-ups (head) office. Here we were expected to wear suits. The novelty made it fun but I could not do that every day.
To an extent, these rules are necessary. It is easy to cooperate when the entire company is in one room (it is also easier to shoot each other with nerf guns). As the organization grows, you need structure to cooperate effectively.
Entwined in workplace culture is the workload expectation. In some organisations, there is pressure to give all of yourself. It doesn’t have to be this way. In my experience, you can find an organisation, of any size, that respects you.
The bank provided a great work-life balance. It was unusual to work overtime. Of course, there were times of high stress. Sometimes you had to buckle down and get the job done. This came with the (unspoken) expectation that you could slack off later.
You can find an organisation, of any size, that respects you
When it came to the actual work, my role was clearly defined. I worked on models and strategies used to automate lending. Rarely did I stray from this objective.
The startup was wilder. Often there was no clear objective. This is the point — to find your objective. As a result, I did work not expected of a typical data scientist. This included data ingestion, warehouse management, developing code and deploying it to production.
Don’t mistake the variety of tasks for an intense workload. I think there is a miss conception that startups will always provide a fast-paced, high-stress environment. This was not by experience. Yes, the work was chaotic but the chaos was contained within reasonable working hours.
At the startup, my net of learning was cast wide. Doing a variety of tasks meant I learned new things every day. From data ingestion to deployment, I touched on every aspect of the model development pipeline. One problem is that I could only catch the shallowest fish. I didn’t develop a deep understanding of any particular task.
As mentioned, at the bank, my role was more focused. I was like a spear fisherman. I swam deep into the pool of model development…
Okay, enough with the fishing metaphors.
What I mean is that I learned a lot in a narrow direction. Initially, I learned at a similar pace to the startup. After a while, my learning slowed. I got good at the role and wasn’t exposed to new problems on a daily basis.
Ad hoc requests also played a part in this. Could you refresh this query you wrote 6 months ago? After a couple of years at the bank, I became the person who knows stuff. A growing amount of my time went towards explaining or refreshing old work. In some ways, I was learning new skills — how to successfully ignore emails.
I didn’t stick around at the start-up long enough for that to happen. I’m not sure if it would have been as bad. There are fewer people who can ask you to do things. The downside to this is there are fewer people to ask when you are stuck.
I often did feel like I was on my own at the start-up. There are no processes or documentation to follow. You need to figure things out for yourself. I’m not saying that the team wasn’t willing to help. They simply didn’t have the answers. The work was new to them too.
The corporate provided more guidance. There were existing, well-documented model development processes. These are not set in stone but they provide guidance on modelling decisions.
A large organisation also means there is a lot of knowledge going around. If I had a problem, chances are someone had encountered it before. People were always willing (after the 2nd or 3rd email) to help. If they didn’t know the answer they would point you toward someone who did.
Personally, this last difference is the most important. At the bank, I worked with a large amount of data. In fact, it was a significant proportion of Irish account data. This meant the data provided insights not only into our customers but into the economy as a whole.
This is what gets me going as a data scientist.
Having that amount of data at the end of your fingertips is enticing. Data warehouses were maintained by entire teams and tables were updated regularly. In some cases, every day. These datasets are a powerful resource I could use to solve problems. They are also incredibly interesting to explore.
In comparison, the start-up was lean on the data side. I did work with large datasets but nothing on the scale of a corporate. The datasets were also static. They couldn’t be used to understand customers in real time.
The upside is that I had full control over the tools I could use to explore this data. Startups do not have legacy code or tools. I could solve a problem any way I wanted. This meant exploring the latest technology and using the best tool for the job.
At the bank… not so much. A downside of having procedures in place is there are existing tools used to implement those procedures. Trying something else often meant jumping through several hoops. Working with big data is fun. Not so much when your tools let you down.
I learned new skills at both the startup and corporate. Yet, the biggest lesson was what I want from my career. That is to do analysis and build models that will have a significant impact on the world. Along with this, I realised I must ask questions that ensure I will do this type of work.
For my next job, I would ask:
- Roughly, how many people/accounts/businesses/etc… are in your dataset?
- How often is the dataset updated?
- What tools do you have available and how much control do I have over the development environment?
The first two questions are to gauge potential impact. With a large amount of real-time data, I would be able to solve some interesting problems. The last question would ensure that my creativity, in solving those problems, was not hindered by old tech. I’m open to working anywhere as long as these criteria are met.
Your criteria will be different. The size of the organisation might be important. Want more flexibility — join a startup. More stability — corporate may be a better option. The important thing is to think about your wants and come up with questions that reflect them.
You may have no idea what you want. I didn’t either when I walked into my first job. Well… besides a desk far away from the guy with a nerf gun. I’m hoping my experience has helped point you in the right direction. Ultimately, you may need to work at a few different places to be sure.
Keep in mind that my experience is also limited. Not all startups and corporates are the same. I haven’t even covered all the ways to work as a data scientist. You can work for an SME or freelance. You can even make money writing about data science on Medium.
The differences in culture, workload, mentorship, learning rate and access to data and tools
Walking into the office looked more like a day at the beach. Shorts and flip-flops. Was this startup culture or just Cape Town? From the CEO to the new hire (me), the entire 8-person team worked from the same room.
During my time there, I briefly consulted for one of South Africa’s telecommunications companies. The contrast was stark. I spent time at the head office in Johannesburg. 100 people in one department and closed shoes — how can I work under these conditions!? The CEO didn’t even know I existed.
Startups and large corporates are on the two extremes of working life. They can offer vastly different work and environments. These differences will ultimately drive your choice of where to work.
I want to share the differences I’ve experienced with a focus on…
- culture,
- workload,
- mentorship,
- learning rate,
- access to data and
- tools used to work with that data.
To end, I’ll touch on how this experience has shaped my aspirations as a data scientist.
At the startup, the environment was relaxed. No dress code (obviously) and the hours were flexible up until a point. If I wanted to arrive late I just worked later. No questions asked. These relaxed rules, extended to the way we interacted. We could talk freely and there was no obvious hierarchy.
For the majority of my corporate experience, I worked for a large bank in Ireland. That environment was less relaxed. Working hours were set and there was a clear hierarchy. Your manager needed to know if you’d be late for work.
There was a dress code but it wasn’t too strict. Jeans and hoodies for the most part. Occasionally, we spent the day at the grown-ups (head) office. Here we were expected to wear suits. The novelty made it fun but I could not do that every day.
To an extent, these rules are necessary. It is easy to cooperate when the entire company is in one room (it is also easier to shoot each other with nerf guns). As the organization grows, you need structure to cooperate effectively.
Entwined in workplace culture is the workload expectation. In some organisations, there is pressure to give all of yourself. It doesn’t have to be this way. In my experience, you can find an organisation, of any size, that respects you.
The bank provided a great work-life balance. It was unusual to work overtime. Of course, there were times of high stress. Sometimes you had to buckle down and get the job done. This came with the (unspoken) expectation that you could slack off later.
You can find an organisation, of any size, that respects you
When it came to the actual work, my role was clearly defined. I worked on models and strategies used to automate lending. Rarely did I stray from this objective.
The startup was wilder. Often there was no clear objective. This is the point — to find your objective. As a result, I did work not expected of a typical data scientist. This included data ingestion, warehouse management, developing code and deploying it to production.
Don’t mistake the variety of tasks for an intense workload. I think there is a miss conception that startups will always provide a fast-paced, high-stress environment. This was not by experience. Yes, the work was chaotic but the chaos was contained within reasonable working hours.
At the startup, my net of learning was cast wide. Doing a variety of tasks meant I learned new things every day. From data ingestion to deployment, I touched on every aspect of the model development pipeline. One problem is that I could only catch the shallowest fish. I didn’t develop a deep understanding of any particular task.
As mentioned, at the bank, my role was more focused. I was like a spear fisherman. I swam deep into the pool of model development…
Okay, enough with the fishing metaphors.
What I mean is that I learned a lot in a narrow direction. Initially, I learned at a similar pace to the startup. After a while, my learning slowed. I got good at the role and wasn’t exposed to new problems on a daily basis.
Ad hoc requests also played a part in this. Could you refresh this query you wrote 6 months ago? After a couple of years at the bank, I became the person who knows stuff. A growing amount of my time went towards explaining or refreshing old work. In some ways, I was learning new skills — how to successfully ignore emails.
I didn’t stick around at the start-up long enough for that to happen. I’m not sure if it would have been as bad. There are fewer people who can ask you to do things. The downside to this is there are fewer people to ask when you are stuck.
I often did feel like I was on my own at the start-up. There are no processes or documentation to follow. You need to figure things out for yourself. I’m not saying that the team wasn’t willing to help. They simply didn’t have the answers. The work was new to them too.
The corporate provided more guidance. There were existing, well-documented model development processes. These are not set in stone but they provide guidance on modelling decisions.
A large organisation also means there is a lot of knowledge going around. If I had a problem, chances are someone had encountered it before. People were always willing (after the 2nd or 3rd email) to help. If they didn’t know the answer they would point you toward someone who did.
Personally, this last difference is the most important. At the bank, I worked with a large amount of data. In fact, it was a significant proportion of Irish account data. This meant the data provided insights not only into our customers but into the economy as a whole.
This is what gets me going as a data scientist.
Having that amount of data at the end of your fingertips is enticing. Data warehouses were maintained by entire teams and tables were updated regularly. In some cases, every day. These datasets are a powerful resource I could use to solve problems. They are also incredibly interesting to explore.
In comparison, the start-up was lean on the data side. I did work with large datasets but nothing on the scale of a corporate. The datasets were also static. They couldn’t be used to understand customers in real time.
The upside is that I had full control over the tools I could use to explore this data. Startups do not have legacy code or tools. I could solve a problem any way I wanted. This meant exploring the latest technology and using the best tool for the job.
At the bank… not so much. A downside of having procedures in place is there are existing tools used to implement those procedures. Trying something else often meant jumping through several hoops. Working with big data is fun. Not so much when your tools let you down.
I learned new skills at both the startup and corporate. Yet, the biggest lesson was what I want from my career. That is to do analysis and build models that will have a significant impact on the world. Along with this, I realised I must ask questions that ensure I will do this type of work.
For my next job, I would ask:
- Roughly, how many people/accounts/businesses/etc… are in your dataset?
- How often is the dataset updated?
- What tools do you have available and how much control do I have over the development environment?
The first two questions are to gauge potential impact. With a large amount of real-time data, I would be able to solve some interesting problems. The last question would ensure that my creativity, in solving those problems, was not hindered by old tech. I’m open to working anywhere as long as these criteria are met.
Your criteria will be different. The size of the organisation might be important. Want more flexibility — join a startup. More stability — corporate may be a better option. The important thing is to think about your wants and come up with questions that reflect them.
You may have no idea what you want. I didn’t either when I walked into my first job. Well… besides a desk far away from the guy with a nerf gun. I’m hoping my experience has helped point you in the right direction. Ultimately, you may need to work at a few different places to be sure.
Keep in mind that my experience is also limited. Not all startups and corporates are the same. I haven’t even covered all the ways to work as a data scientist. You can work for an SME or freelance. You can even make money writing about data science on Medium.
Denial of responsibility! Techno Blender is an automatic aggregator of the all world’s media. In each content, the hyperlink to the primary source is specified. All trademarks belong to their rightful owners, all materials to their authors. If you are the owner of the content and do not want us to publish your materials, please contact us by email – [email protected]. The content will be deleted within 24 hours.