How To Find Good Data Scientists For Your Company

Big data makes a big noise. In fact, it can get so loud that it becomes impossible for an “untrained ear” to decipher the information at hand. A data scientist is someone who can help you control the commotion and utilize the data to your (business) advantage. 

But how do you select the best team that will riddle out the data and provide you with the information that will open up new opportunities? 

Design a smart hiring process

In this particular field, typical hiring interview questions will not suffice. To test whether a potential employee has what it takes to meet your expectations, more “technical” questions are necessary, combined with a practical test. This way you get to see their thought process and how long it takes them to solve the problem they are presented with. 

The problem on the test should be a reflection of the work they are going to do on an everyday basis as a part of their job description. Not only will this show you how well they are likely to perform in the future, but they also get to see whether this job fits their expectations. Unfortunately, it often happens that, even though you discover an ideal candidate, they wrongly assume what their job description entails and decide to leave after a few months, putting you back in the hiring mode. 

Assess their knowledge

Expect the pool of candidates who apply for the position to have diverse backgrounds. While some may have worked on building ML models, others have strong research and analytics skills. 

The trick is to find a person who possesses relevant knowledge for a problem you wish to solve. An all-knowing data scientist is a unicorn, which is why you need to specify right from the start the field of research. This will help you narrow down the number of candidates and find the one who fills the gap in your current team. 

While evaluating the candidates, you need to keep in mind the product that you are developing. Based on your specific requirements, a data scientist should have adequate expertise - the ability to analyze, visualize, scale, etc. In addition, be sure to assess their quantitative and programming skills, and only then consider whether they fit in with the company culture. This way you eliminate the bias and narrow down the list to the most competent few. 

Remember: Not everything is about qualifications

Although PhD and Kaggle participations are a huge plus, it doesn’t mean that they should be mandatory hiring criteria. Carefully analyzing their previous work experience is far more important, since some of the projects they worked on in the past might have provided them with more applicable knowledge than any educational institution. 

In fact, those who only have exceptional educational backgrounds may not turn out to be better candidates than those with a lesser degree and greater practical experience. The latter ones possess what we call “commercial” experience, and can evaluate whether a particular situation requires you to spend hours analyzing data or you can simply deliver a prototype to proceed to the next stage of the project. 

Broaden your search

To hire the perfect candidate, you will sometimes need to expand your search and make it global. Finding data scientists who fit your specific criteria is challenging, but not impossible. To maximize their chances, companies decide to implement an outsourcing business model. It allows them to consider the best possible candidate for the job without having to compromise. A perfect data scientist is not a unicorn, but they might be thousands of miles away. Luckily, these days, that’s no longer an obstacle. 

Include the entire team in the hiring process

Your data scientist will not be working alone. Their job will require them to collaborate with the product development team, which means it is imperative that they get along. Let’s be honest - no one knows better which skill set is necessary to optimize the process than your development team, which is why managers cannot make a confident decision on their own. 

Be the brand everybody wants to work for

Finally, all the above will be in vain if you don’t manage to sell yourself as the dream employer. Data science is a competitive field, and there are still only a handful of truly competent data scientists who will meet your expectations.  

Streamline the hiring process so that the candidates would move fast through the funnel. This will keep them engaged and not have them jump at the next opportunity while you are still weighing your options. Remember to also keep the process interesting to attract the candidates and make them want to learn more about the product they would be working on. 

Familiarize them with the goals and future plans, discuss the possibilities for growth and improvement from the start and you’ll be able to determine which of the applicants is most suitable for the position. In the end, the recruitment process should be designed in such a way that both you and the data scientist sitting across from you can determine whether that really is the job for them.