As a business owner, you may be seeking a data scientist to help you derive value from your company’s data assets. This is a strategic move, one that your competitors are likely making as well, so you want to make this important hire count. In this article, I will cover several reasons why you might find yourself spinning your wheels and encountering frustration during the search process to fill the position of data scientist.
Confusing Data Scientist with Data Engineer
One big stumbling block in hiring a data scientist is confusing this position with a data engineer. I regularly see job ads that confuse these two very distinct job categories. If you’re hiring a data scientist, make sure your job ad doesn’t include skillsets for data engineers who tend to be the designers, builders, and managers of the information or big data infrastructure. Skills required for these areas include: data ingestion; data synchronization; data transformation (ETL); data governance and security; performance optimization; and production orchestration. You could also include experience with distributed computing framework like Hadoop and/or Spark.
A data scientist, on the other hand, should have the following skills: non-technical skills like intellectual curiosity, business acumen, communication skills for data storytelling; academic background in computer science and/or mathematics, statistics; coding in R/Python; data access with SQL; data analysis; data visualization; and machine learning;
Looking for a Unicorn
It could be that you know the differences between the data scientist and data engineer roles, but you prefer to conflate the two and try to find one single individual with all the skillsets. It’s probably unwise to seek out such a “unicorn,” since they’re very hard to find (and could stretch out the time period for your search process), they’re very expensive, and if something happens to the unicorn (I’ll leave this up to your imagination), you could be left high and dry.
A “team” of data professionals, data engineers along with data scientists is preferable. Your pursuit to hire a data scientist will be considerably streamlined if you adjust your expectations.
Need for a Mentor
Another reason why hiring a data scientist might be perplexing is a general lack of understanding of data science. I’ve found this to be particularly true of young start-ups that see their data as a competitive advantage (which is good), but don’t necessarily have the expertise to cultivate this notion. Hiring a data scientist is the first step, but if you don’t understand the field, you may enter a state of hiring paralysis.
Instead of going it alone, you might consider hiring a consultant to guide your path. Hiring a data science consultant to help you hire a data science employee, may sound silly, but it can clear up a lot of misconceptions and confusions. A data scientist will know what to look for, what interview questions to ask, and what skill sets are desirable. You’re likely not to be hoodwinked by charlatans when you have a real data scientist in your back pocket.
You might ask – why not just hire the consultant and be done with it? The answer is you could, but a consultant is a consultant for a reason. Maybe he/she likes that kind of employment relationship, rather than a full-time position. Besides, you might not want to pay a consultant’s hourly fee on a full-time basis.
As a consultant myself, I recently engaged with a start-up company in exactly this type of relationship, and it’s working out great. But instead of working to fill full-time positions, the company is trying to contract with an outside data science team to develop a machine learning solution for a business problem. I’m helping them evaluate a number of outside data science teams, and out of nearly 20 candidates, we’ve narrowed it down to just three. I respect my client’s realization that they don’t have the experience to hire the team themselves. It is a strategic decision that will cut their risks considerably.
In summary, if you’re experiencing frustration with hiring a data scientist, you’re likely experiencing one or more common pitfalls. I’ve outlined several, but there are many more like lacking a budget for data science (a realistic financial commitment is necessary). In this competitive market, you can’t skimp on data science salaries. You also have to consider who the data scientist reports to, IT may not be the appropriate place. Finally, does your company possess a culture that’s compatible with data science; are you analytics-driven? A data scientist will only thrive in an environment where analytics drives decision making.
For organizations experiencing difficulty in securing the talents of a data scientist, you may wish to review some common mistakes others have committed before you proceed any further.
Editor’s note: Want to learn more about how to get your AI initiatives off the ground? Attend the Accelerate AI conference this April 30-May 3 in Boston and learn all about connecting AI to your business today!