What does it take to Lead Data Scientists?

You think it’s hard to hire data scientists? Try hiring a data science leader!

Actually, a lot of people are trying — after all, it’s tough to have a data science team without someone leading that team. And, based on what I’ve seen, a lot of folks aren’t sure what to look for in a data science leader.

Here’s what I believe it takes to lead data scientists effectively:

Being a data scientist.

Not necessarily a great data scientist, but at least a good one. A data science leader needs to be able to distinguish reality from hype — a growing challenge, given the noise in the marketplace. Also, given how hard it is to hire and retain data scientists, a leader needs to establish credibility with the team. Moreover, given the wide variety of people calling themselves data scientists — from management consultants to machine learning engineers — bear in mind that the leader is likely to shape that team in his or her image. So hire a leader aligned with the team you’d like to establish.

Knowing how to hire.

Hiring is a useful skill for all leaders. But it’s especially important in a field like data science that lacks standard credentials or hiring rubrics. A data science leader needs to drive the team’s hiring process, from writing compelling job descriptions to efficiently sourcing candidates and effectively interviewing them. It’s helpful for a leader to have a strong personal brand and a big network, but it far more important to have a principled and rigorous approach to the overall hiring process.

Having organizational awareness.

Organizational awareness is another skill that’s useful for all leaders, but it’s especially important for data science. Depending on the company’s org structure, data science teams often devolve into passive reporting teams or marginalized research labs (cf. “Where should you put your data scientists?”). A data science leader may not be in a position to dictate the company’s org structure. But key part of the his or her job is to align the team’s work with the broader organization while making the team’s valuable — and scarce — skill set.

An open mind about how to contribute.

The scope of data science is a blessing and a curse: it’s not always clear how the team should direct its efforts to create the most value for the organization. Data science can improve customer-facing products and business decisions — and each of these areas are incredibly broad. It’s a leader’s job to keep an open mind about opportunities to contribute while maintaining enough focus for the team to be productive.

Having general leadership skills.

And of course a data science leader needs to have general leadership skills. None of the above skills will help someone who lacks vision, empathy, or integrity. Or someone who can’t communicate effectively.

If you can’t hire them, grow your own.

So, if you thought hiring data scientists was like finding unicorns, data science leaders are more like alicorns. If you’re struggling to hire one, you have my sympathies.

Fortunately, we’re seeing an increasing number of data scientists grow up to become leaders. Perhaps you already have some inside your company.

For more thoughts on how to build effective data science teams, I encourage you to read this article that I co-authored with Jeremy Stanley.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.