The Road To Data Science Maturity - Culture Part 1

The Golden State Warriors (GSW) just won another National Basketball Association (NBA) championship in 2022. This is their 4th championship title in 8 years. They have competed in the finals for a championship a staggering 6 times in the last 8 years. They have won it 4 times. These are astronomical results. Results most NBA teams will be envious of.

Steve Kerr, the Head Coach at GSW has done a tremendous job of building on top of what Mark Jackson did. Mark Jackson instilled the foundations in how the team competed and the type of basketball they play.

Getting to an NBA final takes a whole team effort. From General Managers, Scouts, Agents, Physios, Trainers, Cleaning Staff, Office Staff, Players, Coaches and so many more to name.

This also doesn’t happen overnight. It takes years of hard work, consistency, drive and focus toward one goal. That one goal is winning a championship.

All decisions made about the business, teams, coaches and players are with the mindset of?WINNING.

You are probably wondering at this stage what all of this has got to do with Data Science Maturity and Culture.

What Is Culture?

Culture is how employees’ hearts and stomachs feel about Monday morning on Sunday night. — Bill Marklein

Culture is sometimes quite hard to define but you can sense and feel it. I googled ‘culture’, it’s defined as the ideas, customs and social behaviour of a particular people or society. Companies and organisations have cultures but employees are generally more impacted by their team’s culture which is often set by the leader of the team.

Achieving maturity in data science requires a village. This village will encompass so many different teams who will have their own culture and ways of working. So how do you bring these teams together, to collaborate, solve problems and work as one united force to achieve the business objectives?

The Village Required

The diagram below depicts the village required. The more people you have working in a team, the harder it is to collaborate and share ideas. This becomes even harder when you are trying to get teams to collaborate, solve problems and achieve results together as one team.

This can sometimes be the stumbling block which stops mid-tier data or data science functions from becoming market leaders. This stops them from having a matured data science function. They can’t quite get these teams to work in unison towards one unified vision.

Building Great Culture For Data Teams

To win in this space, or in another language to become a market leader in this space, you must first have a great culture as a collection of teams. Not just in the organisation but on the ground level where the real work gets done.

As a trained mathematician and data scientist, I love having a structure to concepts or thought leadership pieces. Below is the structure I came up with to build a great culture in data science and its ecosystem. This is not a comprehensive list but the rest of the article will delve into the pillars of this structure in more detail. Please read part 2 of this article which talks about the pillars not covered in this article.

Team Dynamics

Two things come up within team dynamics; ways of working and collaboration. This has a lot to do with how the teams are set up. There are a lot of articles out there that talk about team structures and how to best set up technology teams. One thing to consider is that data science, or the delivery of data science, is not always the same as the delivery of software or a platform. Hence, you can’t prescribe all of the engineering principles to data science.

I’m a big fan of getting all the people involved in the delivery of a project working together in squads as part of a tribe. The tribe can have multiple squads that align with the delivery vision. Being aligned to a common product vision is very important which can then be transcribed into a delivery vision.

For example,

Customer Need?— Our customers want to receive recommendations on which product to browse next in an online retail shop based on their browsing history or liked catalogue.

Product Vision?— We want to improve the conversion rate on customers’ next best action on our online retail site. Next click conversion rate.

Delivery Vision?— We want to understand our customer’s next best action. What data do we have available? What analysis can we derive from this data? What solutions or models can we build to help improve this metric?

The Feedback Loop?— Communication and teamwork are required along the whole journey from conception to the delivery of this service. Product leaders should be constantly seeking to understand customers’ needs and engage with the tribe to understand not only how the customer’s needs can be met, but how they can be met and add value to the bottom line or customer experience.

GSW Example

The Golden State Warriors are the best-known team for team basketball. They play unselfishly and it’s so beautiful to watch the game of basketball played the way they do it.

Yes, they have world-class talent in two of the best three-point shooters the game of basketball has ever seen. Steph Curry and Klay Thompson. Their style of play which involves feeding the hot hand along with finding the open man is what sets them apart from their counterparts.

Each member of the team understands their role so well and understands how they play a part in contributing to the team winning basketball games. Their team has a style of play and a DNA that they all bought into and work together to achieve great results.

When you watch their games each team member, even those on the bench is cheering for the team to win. They just look like they are having fun and enjoy working together. It took years to develop their playing style and they lost a lot of games at the start of the journey.

As a leader, be on the lookout for these dynamics in your team. Does this village enjoy working together? Does each team know the role they play in achieving the delivery/product vision? Are the team enjoying the process of working together?

Rewards/Talent Management

A lot of companies still don’t understand the value of their data science or technology teams. Rome was not built in a day, so how do leaders in this space design talent and reward strategies that retain top talent and build upon culture? If you are rehiring your data science team every year or 1.5 years it’s hard to capitalise on any momentum.

I have worked in organisations where the whole data science team have walked out the door in 3 months due to the unwillingness of Senior Management to reward them for their efforts and/or a lack of understanding of the value they have generated.

I like how start-ups and scale-ups encourage continuous learning and development. Not all rewards have to be financial. I’m not advocating for free lunches and pool tables, though they are nice that’s not the reason an employee will stay with an organisation. If you have a strong and healthy culture, employees should enjoy coming to work not because they need to pay their bills but because what they are doing lights a fire in their souls. They want to play a part in the vision coming to life.

Your culture should be so strong that if you hired the right people they would be willing to do what they do for free. Because it aligns with their values, passion and purpose.

These are the people you want to retain in your teams and promote. Not the superstar developer or data scientist that is an a******, that no one wants to work with.

“The real company values, as opposed to the nice-sounding values, are shown by who gets rewarded, promoted or let go” — Netflix Culture Deck

GSW Example

The National Basket Association (NBA) has a yearly draft to bring in new talent. If your team just won the championship you are usually not going to get a high enough pick to bring in the best talent.

This has never stopped GSW to recruit players who seem average on paper and turn them into superstars. A few things come to mind.

When the superstars and veterans on your team are bought into the system of play and the ethos around the team, it’s easier for young rookies or new recruits to get embedded quickly because everyone follows the system of play. This also makes it easier to get value from new recruits a lot sooner.

GSW drafted Steph Curry when a lot of teams weren’t too keen on drafting him due to his ankle injuries. GSW invested heavily into strengthening Steph’s physique to cope with the demands of NBA basketball.

Overall, GSW has done a great job to retain their core superstars Steph Curry, Klay Thompson and Draymond Green. Loyalty has been built both ways. The organisation renewed a contract for Klay Thompson who had an ACL(Anterior Cruciate Ligament) injury. A 5-year extension worth 190 million dollars for a player who would not be contributing to the work on the court for a whole year.

The organisation saw the value in retaining Klay for his previous contributions to the team and for the potential of helping the team win championships again. GSW are a great example of cultivating and nurturing the talent they have at their disposal.

The talent pool in data or data science is growing but top talent is still rare to come across. In addition, a talent that will fit into your culture and be bought into the vision, ethos and ways of working can be even harder to find.

If you find a top talent what is your retention plan? What is your development plan? How do you create an environment for them to grow and learn new skills whilst at the same time contributing to the bottom or top line?

No one leaves a company where they are valued, treated with respect and feel like they are making an impact.

Part 2 of this article will cover the rest of the pillars that lead to a great culture. Be intentional about building your data or data science team culture.?WINNING?teams have healthy and strong cultures. Achieving maturity in data science requires a great team.