- Build an Analytics Foundation of Data Literacy and Data Culture
- Analytics is an Iterative Process and Teams Must be Able to Adapt
- Data Analytics Teams Must Interface Between Business and IT
- A Strong Data Foundation is Necessary for Business Growth
Every organization wants a data analytics team that delivers results. Here’s how to structure a winning data analytics team.
Data analytics careers are growing at an exciting rate.
In fact, companies are so enamored with data analytics that the question they now ask themselves is no longer whether they should invest in analytics, but how much.
The goal of this substantial interest in data analytics programs is for companies to make better decisions as a result of their ability to turn massive amounts of data into something usable and understandable.
Many companies are quick to purchase data analytics technology but don’t give enough thought to the creation of their data team.
Building a centralized data team is even more important than choosing which technology or software to use.
Jeremy Adamson, Director of Intelligence and Analytics at WestJet, is a leader in AI and analytics strategy, and he wrote the book on how to organize data teams: “Minding the Machines: Building and Leading Data Science and Analytics Teams.”
“I hope to contribute in some way to make [data analytics] more easily implemented as well as help practitioners give advice to differentiate and move up,” said Adamson. “Most companies tend to get it wrong in their first iteration.”
Data scientists and data engineers serve as the bridge between management and cutting-edge technology. These professionals use quantitative analysis to forecast consumer trends, gauge fraud risk, and make strategic business decisions. A data scientist must be an expert in business operations and IT; they must be able to tell a story through their data.
“I see data science/analytics as a creative practice. Artisans with very different sets of needs than someone who is strictly IT,” said Adamson. “They need to be part of that solution build. You don’t ask a data scientist to build an attrition model. We ask them, ‘How do we reduce attrition?’ It’s a very close but different question.”
Adamson’s new book comes at the perfect time to address the growing pains companies feel as they progress in their data analytics journey.
“The amount of activity right now around building data teams is incredible,” said Chris Sorensen, CPA, CGA, President of Iteration Insights. “I think Jeremy’s book is well-timed. A lot of organizations are building analytics teams but the question around any team is how well built and balanced it is from a personality and skills perspective. Asking these types of questions at the start offers a more pragmatic approach to team building that avoids growing pains down the line.”
Here are three things that experts like Sorensen and Adamson agree should be considered when building a data team at any organization:
Build an Analytics Foundation of Data Literacy and Data Culture
The data culture in an organization can either help or harm the data team. With a good environment and support from management, they can reach their full potential. They can create more meaningful, actionable insights for the business to then focus on doing what they do best.
But how do organizations find the right people and establish data culture?
“Getting the first person on the team is the hardest,” said Sorensen. “Getting the right person to plant the seed. You have to put some thought into what you’re doing first.”
Adamson agrees that data literacy, culture, and knowledge must be cultivated from the ground up. He said, “Many companies begin their analytics journey looking for ways to cut costs and automate. New products and innovation are where the real value is. For a leader in the practice, to be able to do that you need to create an environment in which it can happen.”
A data culture also allows for increased maturity across the board. Generally, as a data team matures, it becomes more efficient at creating insights and finding patterns. This is because they know how to use their resources effectively and can refine their process over time. Being able to take advantage of big data can put a company ahead of its competitors with the right knowledge and insights, which leads to higher profits.
“Finding good analytics talent is challenging,” said Sorensen. “If you start and build something special from the ground up and invest in people’s careers and give them a place to grow and offer support, they tend to stick around. That’s the type of culture we are trying to create here at Iteration Insights.”
With good processes in place, organizations can continue to grow and scale quickly and easily. Iteration Insights breaks it down by:
“Everybody wants to start with technology,” Sorensen said. “That is 100% the wrong place to start with this stuff. You’ve got to start with people and processes first then work your way up to the technology.”
Analytics is an Iterative Process and Teams Must be Able to Adapt
The two most important factors in analytics are also the most underrated: iteration and adaptability.
It seems like a no-brainer that analytics teams should be iterating on their processes and making changes that improve their understanding of data.
However, in the case of many companies, it becomes clear that this is not happening. It often boils down to a communication issue.
“Collaboration and communication are key,” said Sorensen. “The team needs to be adaptable, iterative, and grow with twists and turns to continuously deliver value in the organization.”
Analytics should be a process where all parts of the organization work together to solve one problem, which is to get insight from data. Teams should never settle for a product or process that does not generate insights.
“Analytics has always been a very iterative process,” said Sorensen. “Questions are a good thing. It means they value your services and keep coming back for more. That’s exactly what you want.”
Data & Analytics Organization Models, Roles, and Responsibilities
Once some of the big-picture constructs of creating a data team are underway, it’s time to start thinking about the actual structure and composition of a data team.
An organization’s data team itself can change and grow as business needs shift. Sorensen laid out a few scenarios of how this may look, depending on the size of the business and its goals.
In general, data teams need frontend and backend people. Data analytics team responsibilities are split up among these two groups:
- Frontend people are good collaborators and communicators that can work with business users.
- Backend people can be more technical, focused on building solutions.
On smaller data teams, people will need to wear multiple hats. If there are only two or three people, one person may have to be the data engineer, the data visualization person, and the requirements gathering person. Larger teams have more specialized roles.
“Teams need a good variety of skills,” said Sorensen. “You need somebody who can interface with the business and also has enough technical acumen to deliver that to the users. Purely IT people will likely miss the mark.”
Data Analytics Teams Must Interface Between Business and IT
The data science department is a critical bridge between business and IT, and a broad range of skills must be represented in order to perform in both IT and business.
“The key is to build a team that interfaces well with the business,” said Adamson. “You can host thinking sessions and leverage good relationships. A creative, positive environment can encourage these ideas to come up naturally and teams can partner to execute on them.”
It needs to be looked at holistically if you want it to be successful.
“The teams themselves have to be good at collaborating and communicating within the team and the business users,” Sorensen said. “We’re not building data environments just for fun. We’re building them for a business user to use and get some value out of.”
A Strong Data Foundation is Necessary for Business Growth
As we’ve seen in past technology waves, all of this excitement will inevitably lead to people calling themselves data scientists without actually knowing what they’re doing. Companies that dedicate resources to establish a strong data culture are best prepared to create meaningful insights.
“Building these great data teams right now will set many organizations up to achieve advanced analytics, which is where everybody wants to go,” Sorensen said. “Everybody wants to jump into advanced analytics but very few people have the foundation to do that successfully.”
A centralized data culture creates a collaborative environment where data scientists, IT teams, and management work together. This allows everyone to have a common understanding of what data is available and how it can be used to drive decisions.
“I cannot believe how many organizations right now are seeing value in building data teams to get a deeper understanding of how well their business is performing,” said Sorensen. “When you think of analytics, most people think of a finance team doing plain vanilla reporting. But it can be a very interactive and engaging thing if you do it right. It’s taking numbers on a page and bringing them to life.”
As more companies continue to incorporate data analytics into their business strategy, standard practices will become the norm. For now, data analytics is still a relatively new practice but professionals like Sorensen and Adamson are actively contributing to industry knowledge and best practices regarding implementation.
“I hope to contribute in some way to make analytics more easily implemented as well as help practitioners give advice to differentiate and move up,” Adamson said. “Understanding the practice and knowing what best in class looks like go a long way to integrate data analytics into executive strategy. It can’t be treated as an afterthought.”