How to Make (and Keep) a Data Scientist Happy
I have seen so many opportunities putting up the data science/machine learning flag for attention, hoping to get numerous highly enthusiastic applicants to apply for a position in which the Machine Learning part is a complete mirage. Even though this approach will attract many people in the short term, it will cost a great deal to the company in the long term as they start losing deceived employees they have invested time, training and resources in. As much as Machine Learning can be a very sexy term to see in a job opportunity, it is a tiny aspect of the data world and most experienced data scientists know it. While the terms might appear less glittery, there is nothing wrong with Data Analytics, BI, Data Quality or Python Development positions as long as they are clearly defined as such.
When setting up a problem statement for a project, I always prefer having too much details and requirements than not enough. How can you satisfy someone without knowing what they need in the first place? I obviously think that I have an idea of what the stakeholders could use, and what could be interesting for the business, but I was surprised several times by demands or requirements that the stakeholders were facing. Having a clear picture helps me deliver the right product and minimize the amount of rework to do and consequently, the amount of frustration to experience. Moreover, if there are doubts on what sort of requirements is needed, stakeholders should feel free to start the discussions with the data science team early, in order for them to set up an accurate problem statement or assignment together.
When we think IT, programming, data science, AI, ML, communication skills do not often come to mind as fundamental skills to have. As mentioned earlier, projects start with goal definition, expectations, needs. Lack of transparency or bad communication can be at the heart of team issues and will grow problems faster than you can imagine. Don’t get me wrong, speaking a lot does not mean clear communication. I know it, I’m hyperactive and Southern French. I can speak for 15 minutes while having said absolutely nothing. Nonetheless, I value teams in which I feel free to express a need for help or advice instead of spending 3 hours losing my marbles while cursing at my Python code.
I thought a lot about combining Clarity and Communication together under one of the Cs as they have so much in common. Honesty and Transparency are some of the valuable aspects needed for both good communication and clarity. However, I care so much about the communication aspects of my job that it needs to have its own C. In any case, honest and transparent leadership can have wonderful consequences on a team and its performance. According to a Forbes article about transparent leaders, “transparency allows relationships to mature faster, as openness can potentially avoid misunderstandings that can fuel unnecessary tension.”
I noticed that in most cases, a few people in a team appreciate the communication aspect of the job more than others. I do not know whether I am more suited than others but I definitely enjoy presenting, discussing, chatting or networking. Recognizing this type of skills in data scientists allows a team or its leader to choose the right person that will be happy to discuss the tasks of organizing meetings with stakeholders or team leaders from other department when they require help.
I value challenges and complex projects a lot.
