Data scientist is said to be the “sexiest job of the 21st century”. There is no clear consensus about the exact role of a data scientist, it will depend on the company, especially if it is a big or a small one. But we can say it has to do with data and its understanding in order to provided some insights to different actors.
Without talking about consultancy which is a particular case, I can see two kinds of data scientist positions:
- Data scientists that come as support. They are not the core of the company, but they help its development by measuring/predicting the product impact on the market or targeting the best audiance for a given product. They help decision makers making the right decision giving advice. Applications are, for instance:
- Build a recommendation system for customer: based on their previous purchases, what are the products they are likely to buy next?
- Test which marketing approach works best through A/B testing.
- Data scientists that make the product. They do not just help the business, they ARE the business in the sense that they build the product that will be sold to customers.
Let’s take the example of a fitness tracker. This device aim is to provide the person wearing it some information about his/her activity, such as number of steps, floor, speed when running… To achieve this objective, the device contains several sensors such as accelerometer, barometer, gyroscope, GPS… But these sensors do not give directly the number of steps. Accelerometer will say “Oh, I moved in that or that direction”. Someone has to interpret these moves and translate such data into the final information the customer want to see, ie how many steps he did. If the algorithm in translation step is not accurate enough, no one will trust the tracker, hence no one will buy it and the company is dead. This is why I say the data scientist is the product maker (even if, in that case, many people have to work together to build the device, not only the algorithm). If interested in this topic, see for instance how fitbit counts steps.
Here is the question: what kind of impact do you want on your business? No the only question to ask yourself before choosing your next position but, in my opinion at least, one of them.