To get a high-performing analytics team producing insights that matter, you need great people, powerful software and a culture of experimentation and innovation. Three simple ingredients, but getting there is far from easy. In this post, I’d like to get you thinking about how to organize for success by building an innovative team that can solve your analytic challenges.

Step one: High-performing people

It all begins with hiring the right people. I recently read How Google Works by Eric Schmidt and Jonathan Rosenberg. This excellent book describes the single most important task for any manager as recruitment. Get this right and you can exploit the herd effect: High performers want to work with other high performers; talent attracts talent.

So, what qualities should you look for when hiring? Deep technical specialization isn’t necessarily at the top of the list. Expertise can be outpaced by changes in technology. The most important qualities to look for are attitude, potential and evidence of continuous improvement and learning.

Data science is a broad, dynamic discipline, so a team has to be flexible. Individuals will need to acquire new skills and take on new roles as the technology landscape and projects shift. Your team needs a growth mindset as described by psychologist Carol Dweck. The fixed mindset is resistant to change; the growth mindset encourages individuals to develop their potential and tackle new challenges -- key attributes for any high-performing analytics team.

Step two: A high-performing analytics platform

If you have the right people, you need to create the right technical environment for their success. What does a high-performing analytics team need to succeed? An analytics environment that’s designed for speed, innovation and autonomy.

The era of having to get it right the first time has been replaced by a fail-fast philosophy, i.e., getting it wrong lots of times and iterating quickly, using a feedback loop to get to the best solution. Why test one analytical model when you could test many, fail fast and pick the best model from experimentation? Why wait for a deployment into the production environment when you could be performing multiple deployments every day? This ‘design for speed’ ethos is very much consistent with the philosophy behind the DevOps movement.

Talented people want to work with technology that helps them push boundaries and gives them the greatest chance of being successful in their role. This often means exploiting new infrastructure or software by ensuring that your analytics platform delivers a modern suite of technologies. Think of it like this – it’s not only your people that need to continuously improve – your analytics platform needs to continuously improve too. You don’t always need the latest and greatest software to be successful, but don’t make your team work in the Dark Ages!

Your platform also needs to allow data scientists to exercise their creativity. They need the autonomy to explore, experiment and challenge throughout the analytics lifecycle. Supporting this autonomous behavior from a platform point of view means ensuring that the team has access to the data that it needs and a portfolio of analytic products and languages.

An analytics platform shouldn't force a particular modus operandi for every data scientist. It should provide a range of interfaces from visual interaction tools and programming interfaces to open APIs that can be called from other languages. This platform flexibility provides the team with the freedom to exploit a range of interaction styles whilst sharing a common analytics environment.

To meet the demands for speed, innovation and autonomy, a high-performing analytics team needs lots of data and the capabilities of a modern, high-performance computing environment.

Step three: A high-performance culture

A high-performance culture is one that evaluates the current status and executes a sustained plan of continuous improvement in a measured way. Some of the traits underpinning this culture include:

• A spirit of innovation and experimentation.
• A focus on results.
• A collaborative environment.
• Continuous learning.
• Embracing failure by learning from mistakes and creating the organizational memory so as not to repeat them.

Another sign of a high-performance culture is support for analytics at the highest-levels of the organization. In this day and age, data and analytics must be at the center of any corporate strategy for success.

Creating this culture is easy, because it's set at the top by anyone who leads a team or is responsible for managing people. A high-performance attitude is contagious because it generates results. If you’re a team leader or a manager, your passion for high performance and drive to ensure that your people fulfill their potential will seep into the team and cascade throughout the organization.

With the right combination of people, technology and culture, you can set your high-performing analytics team on the road to success. Ready to get started? Check out this e-book: Your data scientist hiring guide.