5 Things I Learned as a Data Scientist

If you ask a kid what they want to be when they go grow, is the answer ever “A data scientist?” Well, at least when I was a kid, that was unlikely to be one of the answers. However, when I look back and try to remember when my interest in data started, I can trace it back to my childhood.

My journey to ‘data-land’ started with the enthusiasm for strategy games and developed with sports statistics and online fantasy basketball games. Then, when I was a junior in college, I met contract bridge, a card-game of partnerships which classifies as a mind-sport. ‘The king of the card games’ they call it, or ‘the card game of the kings’ as the likes of Bill Gates and Warren Buffett verify. Since then, almost 10 years now, I’ve been a fan and player of the game. I have been on the Junior National Team a few times and won the U-26 World Championship Title in 2013.

Now that I have brought up that ‘world title’ in my story, I can move on to tell the connection between these hobbies and a data-related career.

In my experience, strategy-based games and data science have two characteristics in common: First is the never-ending competition, and the other is the ongoing quest to make the best use of gathered information. To win in these games, you need to get the data in your reach, then eliminate the misleading or irrelevant data and turn data into information, then use it to decide the best move going forward. Do you see the similarity? This is pretty much why the world is crazy about data these days, we all want to use it to make better decisions. I was happy to discover that one can make a career out of turning data into information. So, I started looking for a job that was an extension of my hobbies. I got lucky and have been working in this field for more than four years now, learning new stuff every day.

In every data-related seminar, you’ll hear about the massive amount of data being generated every day. As we try to turn this bulk data into usable information, the need for professionals rises dramatically which brings us to data scientists, the best job in America according to Glassdoor’s 2018 Job Scores

‘Data Science’ has many aspects, it mainly helps us solve analytically complex problems as well as enabling the creation of real-time, data-driven products and features. Do “People you may know” or “Jobs you may be interested in” ring a bell? Data science is the study of finding information, figuring out what it represents and turning it into a valuable resource to create or grow business and strategies. A data scientist works in this area, there’s not a definitive job description when it comes to this role, but we could simply say that a data scientist is expected to deliver business impact through the information derived from the available data.

Here are five lessons I learned through my experience as data analyst and a data scientist so far:

#1 – Data: Use it or lose it

Data is a messy thing, you will rarely have the perfect data and you should not let this get in your way. An important part of this job is dealing with the imperfections in data. You need to spend time to make sure the data becomes clean enough, starting with the definition of ‘enough’. If you do this right at the beginning, you’ll have a lot less to worry about when you move further and the results will be more reliable.

#2 – Data scientist is not just a data analyst with better technical skills

Yes, a data analyst typically works on structured databases and Business Intelligence (BI) tools, while a data scientist is also expected to build statistical models, leverage machine-learning technology and have programming skills. But this is not the only difference.

A data analyst is generally given the question and uses the question as a guide to find a solution, whereas a data scientist is asked to formulate the questions that will help the business and work through the solution. This requires more than just better technical skills such as a good business vision, strong communication and data visualization skills.

According to IBM,

“A data scientist does not simply collect and report on data, but also looks at it from many angles, determines what it means, then recommends ways to apply the data across an organization’s leadership structure.”

#3 – Accept that as a young data scientist, you’ll suck at certain areas

To be a good data scientist, you’ll need to have programming skills, knowledge of statistics and understanding of the business. Let’s be honest, not many people under 30 are perfect at all these areas and this is normal.

Because there aren’t enough data scientists in the world and this career being a trendy one, some folks start out in different career paths and then acquire skills to move over to data science. So most junior data scientists are either programmers lacking business skills or data analysts lacking programming skills etc.

Once you get that job as a data scientist and are asked to deliver, acting like you know everything won’t help. You’ll need all the help you can get! Make friends with the experienced programmers to find out about the best methods and tools and how to use them, listen to the executives to get to a better understanding of the business, what matters most and what problems keep them awake at night, and you may also need some help from statisticians to build better models. That means besides the main skills, you’ll also need very good communication and teamwork skills, both to deliver the best results and to improve yourself. This is an overlooked skill and a great way to differentiate yourself.

#4 – Getting results is just half of the job. Next, tell a story

Getting results is very important of course, but interpreting and presenting them are just as important. This is where Data Visualization joins the party. Time is precious, so today’s decision makers don’t have time to scan pages and pages of information and decipher what is going on with the business, so you need to find a pictorial or graphical format to let information shine and tell its story. Data visualization helps decision makers to see analytics presented visually, so they can grasp difficult concepts or identify new patterns.

A well-designed user interface, possibly a dashboard, is the most effective and satisfying way to tell a business story with numbers or make a point.

#5 – Curiosity is a must and you’d better have fun

Whatever we do as a data scientist, we should understand what we’re doing and the reason we’re doing it. If you haven’t asked yourself “Why?” or “So what?” for a whole day, you’re doing something wrong.

Most of the time, you’ll be answering the questions that weren’t answered properly or even weren’t asked before. It rarely will be an easy path to victory, get ready for the ups and downs. You’ll make many wrong decisions on the way, you’ll try various methods and fail and there are always others to try, so don’t give up.  Just keep the big picture in mind and stick with it!

Never stop being curious and find ways to have fun throughout the bitter parts of your journey. Curiosity won’t kill you – unless you are a cat and even then you have multiple lives, so carry on and explore!

Leave a Reply

Your email address will not be published. Required fields are marked *