Should you dedicate your analytics skills in a startup or in a big firm? This question has not only confused many freshers but also a lot of experienced data scientists. Although the overall job profile of a data scientist in both kinds of organisations does not differ much, the work culture of data science teams in startups differs a lot from that in large companies. Apart from salaries, how the organisation is shaping you for your career ahead is more important.
Following are some of the ways in which working in both kinds of organisations is different:
In a large company, a data scientist works on the same problem for several months, maybe even years. He/she only need to concentrate on the modelling of the data, and the mathematics or the statistics involved that goes into the model, because the employees of the other are working on the certain specified jobs needed like engineering the data ingest, storage, making pipeline and production deployment. They make you learn a lot in terms of programming and data warehousing. As a result, he becomes an expert in one particular domain.
Startups, on the other hand, make the data scientist’s work uncertain. They work on multiple problems which can be of different domains, at the same time. Their work is not well-defined and they may not be always aware of what work they have to deal with on the day, beforehand. Although this uncertainty makes the work a bit choppy, it gives them the opportunity to build the analytics platform from scratch. He is supposed to be an all-rounder and be an expert in many fields, to successfully contribute to the team.
In big companies, many employees have handled the data, which may or may not suit your understanding. But in startups, only you operate on the data and you get to decide what is to be done with it. If you want to work on something very narrow and specific, one thing at a time, big companies are better. If you want an opportunity to showcase your whole skillset, startups are better.
The skills required for data scientists in both kinds of organisations do not differ much. But in large companies you will automatically be mastering the language that you work in, be it Python, R or Matlab.
Whereas in startups, since the work is not well defined, you may end up using more than one language and not master one particular language, but have a basic knowledge of all. They do most of their work in SQL or spreadsheets.
In large companies, since it is a well-established company with a large number of employees, it has more data and generally will have to use Hadoop.
In small companies, especially the ones trying to find their product market fit, Hadoop is not necessarily required, and they may use PostgreSQL or Vertica.
Since the work is not concretely defined in startups, the data scientist himself mostly identifies the exact problem to work on because he might be the only data scientist in the company. He has to make the other department employees like those in marketing sales and product, understand the problem.
In large companies, however, there will be a boss who understands the problem correctly and defines it for you. Generally, there will also be other data scientists to help you out understanding your task.
Startups have no well-established mentorship and an established workflow. Big companies have a specific way to deal with every project and each person is assigned to specific tasks, making the workflow easy.
Regarding what freshers or professionals usually look for when joining a company, Diego Klabjan, director of the Master of Science in Analytics programme at the Northwestern University and a partner at Opex Analytics had said, “Some of them go start working for bigger companies, and then they find out that there’s not much innovation in those large companies, so they are essentially doing [business intelligence] stuff, but their passion is really data science.”