How to Develop a Career in Data Science
How to Develop a Career in Data Science
Millennium Consulting has a strong track record in sourcing and on-boarding highly qualified technology professionals. Data science is an area that has seen significant growth over the past few years. Given our experience and reputation in this field, we are often contacted by people looking for advice on how to develop a career as a Data Scientist.
To help answer the most common questions we are asked, we interviewed our Head of Recruitment, Brendan Shaw. Brendan has 15 years’ experience sourcing experienced talent predominantly across finance and technology. He has worked with organisations ranging from small commercial clients to large corporate multi-nationals.
What is data science?
Most people think data science is all about machine learning, AI and deciphering codes, however, this isn’t the case most of the time. In reality, data science is more about using data for creating the maximum positive impact for a company or organisation. Data science is a combination of computer sciences and data mining, which is defined by the Journal of Data Science as ‘almost everything that has something to do with data: collecting, analysing and modelling.
From a target marketing perspective, the job of a data scientist is to gain insights on the products, craft product recommendations and use data to improve existing products. Making complicated models and data visualisations come way later in the list of priorities. Market research is a key aspect that needs the work of a data scientist. A data scientist can work through an extensive range of data and find the factors that might improve the sales, or even forecast the demands and expected revenue of a certain product. How people react to certain advertisements on social media can be recorded and then interpreted to forecast the demand or sales that the product is expected to generate. Successful target marketing is a key reason for which companies often rely on data scientists these days.
What’s the ideal path for someone who wants to get a job in data science?
The first thing I would suggest is to study statistics or computer science. Try to do as many internships as possible even if they don’t pay you. Working on real life projects is more important that what you learn through study. If when you’ve graduated and feel you don’t have enough experience, attend training camps or get a job within a company that allows you to improve your skills in data science.
Do you need a bachelor’s degree?
In theory you don’t, and many companies will say you don’t need a degree. In reality though, I haven’t interviewed or come across any data scientist that don’t have a at least a bachelor’s degree. I often come across software engineers that don’t have degrees because in that role you learn more of the practical use of programming language, however, for data scientists, you have a lot of mathematics involved. Without having studied a bachelor’s degree, it’s hard to fully comprehend the science behind it.
While companies with more resources can have a better system in terms of work distribution and employ more people specialising in each aspect of data science, they generally prefer having Master’s or PhD degree holders on board for more advanced and technical aspects, such as jobs involving AI and rigorous machine learning.
I see a lot of candidates with PHDs and masters but if you have experience, honestly, the qualifications don’t matter as much.
What is the best bachelor’s, masters and PHDs if you are wanting to build a career as a data scientist?
If I had to choose just one, it would be statistics. As an all-encompassing subject it’s perfect. I’m a believer in education, formal and informal. Additional subjects that are useful include computer science, software engineering, economics, statistics, mathematics and the sciences.
If you’re thinking of gaining a masters or PHD in data science, analytics, or any other field, because you love it, and you have the time and money to spare, that’s great. But if your interest in analytics is driven by the desire for a job that offers better rewards than you’re getting now, look for alternative paths to make that career move. There are vast career opportunities in this sector for people with different skill levels.
What are the best programming languages to learn?
SQL is considered to be the easiest language for building up the basic infrastructure of a database. It is widely used by people employed in jobs that involve business intelligence. Knowledge in Python, R, Excel and Tableau is also important in order to develop better skills in manipulating, visualising and interpreting data.
What matters more than technical skills is the ability to communicate and interpret the data in a better and easier way. In the end, it really doesn’t matter which software you use as long as you are fast and effective in communicating and interpreting the findings.
Would you advise candidates attend data science bootcamps?
Absolutely. Making the transition into data science can be difficult if you’ve never worked in the industry before. Bootcamps can offer theoretical and practical applications using commercial data science problems from a range of different companies.
You can also apply to companies to work as an Intern Data Scientist. This can help you prepare yourself for the real challenges you might face while working as a professional data scientist. If you don’t have that option but have academic qualifications, bootcamps offer the opportunity to gain practical experience.
Are there any additional skills candidates should work on?
Developing strong communication and presentation skills is really important. Communication is crucial no matter which sector you work in. Within data science, the entire point is to communicate and interpret the findings that come through rigorous analysis. Improving your writing and Power Point skills is important, as good presentation skills will ready help your career.
Is there a high demand for Data Scientist in the market?
Absolutely. Data science is growing much faster than anybody anticipated, with a large amount of data scientists needed. The industry is still trying to work out how to evaluate data scientist, in terms of who has the potential to excel and the perfect characteristics to focus on.
Would you classify machine learning is a subsection of data science?
Machine learning is more focused on algorithms whereas data science focuses on data and insights, paired with algorithms to get the final results. I would therefore personally classify them as two different roles, but others would disagree.
How can candidates gain a competitive edge when applying for internships or their first job?
When applying for any job you need to demonstrate that you have the potential to be the best in that category. If you are applying to be a software engineer, try to demonstrate that you have the potential to be the best software engineer. If you are applying to be a data scientist, try to demonstrate that you have the potential to be the best data scientist.
This isn’t easy as data science is a domain that people need to understand and are still trying to figure out. Data scientist should love data and should be passionate about data. You should have the capability to see data and gain unique insights from it.
Data scientists spend around 90% of their time finding, extracting, cleaning and transforming data and only 10% of the time pairing it with algorithms to get the final results. A data scientist should have the patience needed to clean the data and ability to debug it.
These are all characteristic employers look for and areas you should try and build upon and demonstrate through the application process.