- Building a team starts with business knowledge. Remember that the first word in Business Intelligence is business.
- While having technical skills are important, the soft skills and process of how you work can be vital to the success of a data team.
- Avoid looking for unicorn individuals. Instead, look to build a unicorn team. Unicorns do not exist, and even if you did find one, it is not a good idea to have all your eggs in one basket.
An analytics team is a group of skilled people who work together to solve business problems. Well-functioning analytics teams are essential to delivering business objectives at a high pace and on a sustainable basis.
Understanding the target state composition of a well-functioning team is vital. Determining what gaps to fill (if any) maximizes throughput and ensures sustainable delivery.
Whether your organization is starting its first analytics team or building out an existing one, there likely isn’t a checklist to follow to “build the perfect team”.
I see far too many groups being thrown together without the right skills mix, which hinders their ability to deliver at all or at the least, consistently. Most organizations will benefit from starting small and iterating over time as business needs change.
The demand for analysts continues to grow as more companies realize the benefits of quality data and what it can bring to an organization. According to McKinsey, analytics-driven organizations are 23 times more likely to acquire new customers and nine times more likely to surpass their competitors in customer loyalty.
In this article, I’ll show you how to structure your team so they can deliver more, faster, and with high quality on a sustainable basis.
- Top 3 points to consider when building an analytics team
- Analytics team roles
- Final words and advice
- Analytics team FAQs
Top 3 Points to Consider when Building an Analytics Team
Building a dream analytics team won’t happen overnight, and I encourage you to build out your own roadmap over time. But there are some commonalities I’ve seen that can help streamline the process for your organization.
It all starts with finding your first team member. This person should understand a little bit of everything, emphasizing business knowledge over technology (more on this later).
Here are my top 3 guiding principles that I’ve seen contribute to building successful analytics teams:
- Building a team needs to start with business knowledge
- The first word in Business Intelligence is business
- Avoid looking for unicorn individuals and aim to build a unicorn team
Tip: Building a team needs to start with business knowledge
IT has long had to deal with a misunderstanding of what skills and roles are required, and data teams, which are a subset of IT, are no different.
My parents still ask me if I can help them with hardware problems, and I haven’t the slightest idea how to address them as I am a data guy. Ironically, this type of understanding of what skills are needed for data extends into many organizations.
Instead of starting your analytics team with IT, consider hiring a business-savvy analyst with skills such as:
- Ability to wear multiple hats to adapt as the team scales
- Understands the business
- Capable of collaborating throughout the organization
- Technical ability to navigate programs like Tableau
- Appreciates consulting help
Reminder: The first word in Business Intelligence is business
In this blog, we will briefly discuss the roles to build and run a highly functioning team.
As a leader, it is essential to remember that these are roles and that each member on a team may end up filling multiple. This is especially true in smaller organizations or organizations that have not yet invested in their data teams.
Building a high-performing team is no easy feat and will often take years to achieve.
Understanding the target state is the first step. Compare and contrast this to major sports teams – whereby some with deep pockets can just handpick and buy their team, whereas some need to develop homegrown talent for years.
Building a data team is no different. With more companies leaning towards the slower long-term build and inevitable “patience” ripple effect to the fan base or business users in the case of enterprises.
Advice: Avoid looking for unicorn individuals and aim to build a unicorn team
Too many HR people and team builders want to find that one person to solve all your problems. However, a better solution to your business problems is to develop a well-organized team. This helps not only with the actual work your team is doing but the lasting power of your team as well.
Not only are unicorn people far and few between, or often don’t even exist, but if your team is relying on one unicorn person and they leave, the team will suffer. My recommendation: hire a unicorn team, don’t bother looking for unicorn individuals.
Analytics Team Roles
At a high level, an Analytics team will engage in four different types of work:
- New project development
- Production support
- Ad-hoc query handling
Consider these types of work when determining team size and roles, as each type of work is handled with a different skill set. For now, let’s focus our attention on new project development as it covers a broad spectrum of roles and skills. So, what are those roles?
There is no one-size-fits-all approach to establishing roles for your team, but I can tell you one thing: you cannot build a highly successful data team from Data Scientists alone.
At a high level there are three major groups of roles:
Leadership roles work in the front end of projects and mainly focus on the influence and idea creation for the team. Leaders ensure the project stays on task and deliver information to key players to increase efficiency.
Successful leaders possess a variety of soft skills including communication skills, teamwork skills, motivational skills, analytical skills, problem-solving skills, and conflict resolution.
Leaders and Sponsors
- Not involved daily, but they do influence the team.
- Typically help determine priorities and scope of individual projects or programs
- Often, they work to secure budgets as well
Analytics Director/Manager or Program Manager
- Helps define delivery methodologies, internal process, staffing, and architecture for the overall program
- Ensures that all functions are highly coordinated
- Oversees project orchestration
- Help manage individual projects
- Charts progress
- Manage project issues
- Maintains budget, scope, and upwards communication
- Depending on project size, it might be a part-time role
Business roles are also front-end facing roles and are essential for project success. Business experts are at the heart of any analytics team, and this is where you can really bring your data to life.
Soft skills that business roles include communication, problem-solving, research, attention to detail, and collaboration.
Subject Matter Experts (SME)
- Understand the business process, terminology, stakeholders
- May have done some reporting on their own using Excel or other tools
- May also understand where to find data in the systems and can help interpret
- Consider them a data navigator/your guide
- Gathers and manages requirements
- Helps ensure that they are understood and testable
- Not usually technical in nature
- Liaison between the business and technical team
- Works to help understand business needs
- Job duties/activities may overlap with the SME/PM
Now, it’s time to talk about technical roles. Once you’ve established a strong business side to your team, you may start to hire technical powerhouses to continue the work.
- Where is this all going long term? Help to set a consistent approach to design and delivery
- Manages the design of overall technology perspective
- Keeps program from having disjointed and unsupportable products
- Thinks about systems integration points
- Review process
- Standards > Design
- Influences technology choices to ensure that everyone does not go off and do their own tools and languages. Keep things knit together.
- Familiar with ETL and ELT principles
- Familiar with modern EDW design
- Data movement, transformation, cleansing, and storage
- Typically, understand programming languages such as SQL and Python
- Design targets such as Data Lake etc.
- Helps model optimal storage structure for the data team
- Star Schemas
- Modern DW structures
- Sometimes part of the architect role
- Provides an insightful user experience when consuming a dashboard
- Chooses the right visuals for the right information to ensure accurate delivery of information
- Adds context to the visuals to accelerate comprehension of the information
- Arranges different visuals to a story plot for streamlined execution
- UI and UX skills are beneficial
- Although most organizations do not have this role, I have it here as it is a significant role that can significantly improve overall delivery quality and throughput.
- Can help gather and align metadata (crosses over into BA role)
- Important on larger projects and programs
- On many projects, the BA fills the QA role
At a very high level, these are the main roles that you will encounter most frequently. In larger organizations, you will start to see much more specialized functions such as Metadata Manager, Corporate Librarian, Data Scientists, Data Stewards, or Data Governance.
Final Words and Advice
When reading articles online around tips for building your team, my one main criticism is that most just launch straight into technology as skills that are needed on the data team. Many articles fail to give the proper attention to People (soft skills) and Process (ability to deliver).
Before you start hiring for whatever role you are trying to fill, think about what skills you need for that role. This is important because the market is flooded with people who have hard skills but still might not have soft skills.
I have noticed that many managers are so focused on finding people with hard skills that they will overlook the importance of soft skills.
For example, if you are hiring an analyst, don’t just look for someone who knows SQL or has experience with R, but also someone who has great communication skills. Don’t hire someone just because she knows SQL but he can’t explain her findings in an understandable way.
- The term data scientist is grossly overused and does a disservice to the success of analytics projects as it tries to oversimplify the process. Do not fall into the trap of “just hire a data scientist, and you are done”.
- Avoid looking for unicorn individuals and look to build a unicorn team. Unicorns do not exist, and even if you did find one, it is not a good idea to have all your eggs in one basket.
- Job postings can be beyond ridiculous in terms of what one individual can perform. In the recruiting functions defense, analytics team leaders need to help create realistic postings by collaborating with HR.
- And finally, remember that the role of the analytics team as a whole is to assist the business to better understand and manage performance. Some teams often forget who their customers are and why the group exists in the first place.
Many start-ups build a data team that includes data engineers, statisticians, and computer scientists.
While having these technical skills is important, the soft skills and the process of how you work can be vital to the success of a data team. And if you are missing one of these roles, you can always hire, or engage a consulting firm.
Analytics Team FAQs
What is a data scientist?
A data scientist is responsible for developing analytics solutions that solve business problems.
Data scientists use technical skills coupled with various soft skills to collaborate and communicate within a data team to build models and develop predictive algorithms that allow companies to target customers and produce repeatable business outcomes more effectively.
Why do I need a team? I have myself and Excel.
Developing a successful analytics team takes time and effort, but the right members of your team can help tremendously. Having proper structure in place will not only ensure that your team is successful, but that you are successful as well.
What are some of the characteristics of a good data analyst?
A good data analyst should be curious, collaborative, and communicative, a good storyteller, consider the big picture, and be willing to fight the good fight. They do not jump to conclusions, are not afraid of technology, and know to let the data have a voice.