Analytics & Big Data
Nov 05, 2018

The six traits of winning data science teams

Key factors: the right mix of people who also serve as change champions

In an increasingly digital world, most enterprises amass huge volumes of data. As they become smarter, forward-thinking businesses are using that data to predict and proactively respond to customer behaviors and market trends to deliver real business impact.

Leading firms also appreciate that data science is an open and adaptive field. They understand that its practitioners incubate and adopt new technologies that broaden horizons. So they're creating data science teams.

That raises the question: What makes a good data science team? Talent and skill are crucial, but so is a culture that cultivates curiosity, passion, and drive for the art of problem solving. Here are six ways to help your data science team unleash its potential.

  1. Embrace a mix of skill sets: You breathe life into your business when your team unites technical expertise with robust domain knowledge. It's critical to include people with different business backgrounds and a mix of technology and functional skills. You need to combine machine-learning experts and mathematicians who are skilled in statistical analysis and can understand and apply algorithms to data. A good team also has data engineers who can build data pipelines for analysis and functional gurus – domain experts – who can work through the complexities and challenges of the business.
  2. Prioritize: Once your team is in place, establish an operating model. Just as with preparing a gourmet dinner, you should break it down into workable components. Establish your criteria for success and how you measure it – then communicate that to the team on an ongoing basis. Having clearly defined business priorities helps you demonstrate value. When you regularly share insights with stakeholders and consumers, you can see where roadblocks are holding up the works so you get all your dishes on the table on time.
  3. Get experimental and scale new horizons: In data science, every project begins with a problem and a set of hypotheses. And since you don't start with a real road map, you have ample scope for experimenting and innovating. Questioning minds that thrive on what ifs and why nots are a huge asset in this context. Curiosity is the mother of innovation.
  4. Advocate openness to change: The goal is to make every facet of your enterprise more productive and meaningful, so your team must promote an analytics mindset. Insights from advanced analytics come easy. But those insights won't amount to much unless your enterprise acts on them, so you need to empower the people working in every core business and function to take advantage of these insights. This requires change management, which is not a simple task. After all, it's generally challenging to persuade workers to use new tools, especially if they initially find them confusing. But it's worth the effort. You don't want a work environment that still concentrates on discrete tasks or activities. You're striving to create an integrated, analytics-driven culture – one that is keenly focused on solving overarching business issues.
  5.  Keep IT – and all other business teams – in the loop: Data science experts understand that cross-team collaboration is essential for achieving enterprise-wide business goals. That's why healthy collaboration between analytics and business teams – especially IT – is critical to your overall success. What's more, alliances are more likely to flourish when other teams and functions see the positive impact that data science has on their day-to-day activities.
  6. Accurate data means better analysis: Data quality and availability are crucial – you might even say they're the cornerstone of every organization. After all, your analytics are only as good as the quality and precision of the data you have access to. Robust data accuracy starts at collection, then carries through to data entry, cleansing, and sharing. And as data plays an increasingly vital role for businesses today, many firms are creating new roles and responsibilities to govern it. They recognize that quality data is an essential ingredient for the kind of data analytics that leads to insights and drive business goals.

About the author

Sreekanth Menon

Sreekanth Menon

Practice Leader Data Science, Digital and Analytics

Follow Sreekanth Menon on LinkedIn