Analytics team structures and how they impact you as a Data IC

Where you sit as a Data IC matters

Org structure impacts many things such as day to day work, career growth, learning and development and overall morale.

Having worked in the Data field across all levels from Analyst to Management over 15 years, here are the 3 most common structures adopted by Analytics teams that I have been a part of - Centralized, Embedded and Hybrid. Here you will find what each of them are and what they mean for the Data IC.

Centralised Teams

Under a Centralised structure, there is a single Data & Analytics function that serves the needs of the entire company and all Data roles sit in the one team. They may be split by Analytics function (Data Engineering, Science, Enablement etc) or by the business unit they are supporting, but decision making and management is determined by Leaders within the Centralised team. This structure can be more common in smaller companies and startups because of the low maturity in other areas of the company.

Pros:

• Peer group: The team is made up entirely of Analytics and Data people. This creates a sense of camaraderie and community since everyone has a similar skill set, background and challenges.

• Governance and standardisation: Because everyone’s in the one team, it’s easier to establish standards, ways of working & operating discipline within the team.

• Exposure: Under this structure, there is no single dedication towards a specific business unit. This means work opportunities will develop in multiple areas which is good for variety in both the nature of work and stakeholder groups that are engaged with.

• Career pathway: The functions in this team have clear data-focused pathways under Analytics leadership. This pathway can means progression is a natural to Senior, Lead, Principal and management.

Cons:

• Detached from business: This structure means Analysts are a step removed from the “why” behind the work. This makes it hard to develop any domain business knowledge because this means the team is also a step removed from where the value is generated.

• Reactive: There is less involvement in business unit specific strategy. The share of voice in prioritisation is diminished and incoming work has the perception of being viewed as reactive to stakeholder needs.

• Pivoting: Because of competing priorities, juggling this across different businesses can create confusion as to what is most work should be prioritised.

• Can be a bottleneck: The high demand for Data work from a variety of different areas means there are more delays and more pressure.

Best for: ICs who value technical growth, collaboration with other Analysts and a variety of different project types.

Embedded Teams

Each department (Marketing, Sales, HR, Operations, Finance etc.) has its own Analyst or Analytics function which is fully embedded in the team. This means Analysts are spread out in different business units. This structure exists in organisations where there is a greater need for access to their own Data personnel.

Pros:

• Business alignment: The business domain is well understood along with data needs and how it makes a tangible impact.

• Focused work: Minimal context switching with direct time spent in the business unit.

• Stakeholder trust: Collaboration is easier due to greater accessible to the Data team.

• Proximity to Value: Working as part of an embedded team means Data ICs can be linked more directly with business outcomes.

Con’s:

• Lack of peer support: It can feel isolating to be the only Data person in the room which is often the case under an Embedded structure.

• Data Extracts: The nature of work may be stale and only limited to data pulls and requests with nothing innovative due to lack of data maturity and understandin of other Analytical capabilities in the business unit.

• Tool/process drift: No shared standards across teams, therefore every Analyst within the company may operate differently and use a variation of different tools, coding standards and organisation of resources.

• Uneven workload: For Embedded structures, each team has very few Analysts so the majority of data work tends to fall to too few.

Best for: IC’s who want ownership, domain expertise and be closer to the business value.

Hybrid Teams

This utilises a combination of both Embedded and Centralised structures and can appear in a number of different variations. It could involve a fully scaled up Centralised team or a trimmed down version of this. Within each business unit there may also be single Analyst or a team of them. Either way, both a Centralised and Embedded teams co-exist together. This structure is most common in mature organisations because there is a clear recognition that Data & Analytics functions are needed in every team. At the same time, a Centralised team is formed to coordinate company wide Data initiatives such as governance, literacy and enablement.

Pros:

• Good Balance: Business proximity + analytics community = best of both.

• Support and mentorship: Still have access to data leadership and peers.

• Career growth: Promotion pathways and skills coaching still exist.

• Cross-functional exposure: More opportunities for big-picture projects and collaborations amongst different Embedded and Central teams.

Cons:

• Dual reporting: Conflicting or parralel work can start to arise between Embedded and Central teams.

• More meetings: Forums across both Embedded and Central teams can double up creating multiple meetings.

• Needs maturity: If not managed well, you can get the worst of both worlds.

Best for: ICs who want impact and visibility and yet a strong analytics career path.

So what’s the best structure?

It depends on what stage you’re at and what you value:

• Want to progress up the technical tree and learn from other Analysts? Centralised

• Want to own your own space and be closer to the business? Embedded.

• Want a balance between technical career growth and business impact? Push for a hybrid.

Did I miss out on any? What is your preference?

Let me know in the comments.