If your business has grown past the concept phase, you’ll benefit from regular analytics reporting. Early on, chances are you’ll be able to get a lot out of the built-in analytics of many applications. Eventually, however, your business questions will evolve, and you’ll need more detailed analytics. If you are not familiar with data analytics tools and techniques, you’ll want an analytics team. This post discusses three critical decisions when starting your analytics team.
What Is Your Goal for Starting an Analytics Team?
Fortunately, if you plan it well, your analytics team will evolve with your business, providing support commensurate with your needs. You can read more about the planning process and how to avoid the perils of bolt-on analytics here.
When starting an analytics unit, your most important consideration is understanding the use cases you want to focus on. By use cases, I mean the specific business problems you want analytics to work on.
I recommend you schedule 30 minutes to brainstorm use case scenarios for your analytic team. Be specific about the business problems data sources you want the unit to work on. As you consider the use cases, include any regular reporting you’ll want generated.
Remember, hiring one or more analysts full-time will require enough work for their time. Otherwise, you may want to look for a contractor to provide part-time analytics. If you’re unsure about the time your use cases will require, find an analyst to discuss it with.
Remember that data cleaning and preparation time are always involved in analytics. Whatever time estimate you come up with, double it if data cleaning and prep are not included.
Once you have your use cases and time estimates, you’ll have a solid foundation for structuring the team.
What Will You Invest in The Team?
You need resources to start your analytics team, including time, money, and manpower. Your second critical decision is how much you will invest in each.
The largest time investment in building an analytics team will be recruiting and hiring your staff. Plan on this process taking at least two to three months to complete for a single hire, including:
- Scoping the position(s)
- Writing the job descriptions
- Writing the job postings
- Collecting candidate resumes
- Reviewing and interviewing candidates
- Negotiations (depending on the level of staff hired)
- Onboarding
You will need additional time investments to create your analytics team. Your initial hires might be able to assist with these to reduce your time burden. You must review and select an analytics platform that will work with the rest of your tech stack. Additionally, you will need to develop a plan to integrate analytics into your existing business processes.
Your monetary investments will be determined by your staff’s experience level and time allocations. You’ll need to account for your staff’s salary, benefits, and computers. You’ll also need to account for expenditures on software and any cloud computing/storage.
While manpower resources can easily be included under time and money, I like to think of these separately. I consider the manpower needs in various roles to help confirm my time and money considerations.
Human resources need to develop and recruit for the positions you need, whether as employees or contracted labor. IT needs to acquire, set up, and integrate your software platforms and capabilities. Accounting and finance need to identify and allocate funding. Finally, leadership needs to direct and manage the process: governance, change management, process requirements, and integration.
What Will Be the Organizational Structure?
Your third critical decision is about how to integrate analytics into your organizational structure. If your team is small, the best placement of analytics may be apparent. If your team has grown, you will need to think more about where to place analytics.
The key structural aspect is whether you want analytics to be centralized, decentralized, or matrixed. Most organizations will begin with a centralized structure but may move to a decentralized or matrixed structure as workloads expand.
A centralized structure has analytics housed inside its own business unit, responding to requests from other units. A decentralized structure places analysts across the organization, assigned to units performing analytic work. A matrixed structure combines the two approaches. Analysts from a central analytics unit are assigned to work with other business units.
A centralized structure will make it easier to manage all analytics and create uniformity of processes. In contrast, a decentralized structure allows greater specialization as needed for specific departments. The matrix structure attempts to blend the two but can be more difficult for you to manage.
With a small enough team, you’ll probably want analytics reporting directly to leadership. Depending on the kinds of analytic projects to be done, your initial analytic hire might be your analytic lead. However, if your analyses are not overly complicated, you might be able to hire a lower-level analyst to start.
Conclusion
Leaders working to stand up new business units have no shortage of decisions to make. Starting your analytics team is no different. However, making the three critical decisions in this article will set you up for a successful plan.
