Who Should Own the Business Intelligence Team – IT or Business

Problem

It’s a long-standing question: What corporate facet should the Business Intelligence (BI) team report to? The Business or IT?

This article will probe the centralized versus decentralized BI ownership conundrum. We will provide insights and suggestions on how the BI team can provide deep value regardless of who manages the group.

Solution

Analysts, Engineers, and Developers comprise a traditional BI team. Each bringing a unique set of skills, responsibilities, and value to the team and organization. All these BI roles bring strong analytical skills, such as SQL and Python, to provide data insights, create reports and dashboards, and communicate findings to stakeholders. All roles bridge the gap between data and the business, ensuring that insights are relevant and actionable.

However, BI Developers and Engineers bring advanced skillsets. These include data modeling, ELT processes, database management, and data warehousing to ensure data quality and integrity. They thrive in a programming/software development environment, have BI tool expertise, and understand report platform administration. So, the learning curve for an analyst to transition to a BI Developer/Engineer can be steep. The two personas should not be confused as having the same set of skills. BI Developers and Engineers typically thrive in a centralized IT org. BI Analysts can be effective in an Analytics org within the business.

The Age-Old Dilemma

The age-old dilemma of whether the business or IT should manage the BI team continues in the 21st century. Historically, BI has resided within the IT org for several reasons:

  • Standardization: Centralized control ensures standardization across the organization for governance, security, compliance, and general report administration.
  • Expertise: To manage complex BI tools and infrastructure you need a high degree of technical expertise.

Centralized BI teams can enforce the standardization of data models, tools, and processes, leading to consistency. The following illustration shows a typical organizational structure for a BI Center of Excellence (CoE) team reporting to IT. This shows a strong connection to other IT peers in the Data Warehouse and Data Engineering CoE.

BI Team with IT - org structure showing Bi as part of IT

Why Would Business Want Control the BI Team?

While there are several benefits to this structure, there are reasons why the business may want more control and ownership over BI which are important to consider.

  • Request Delays: IT may have longer turnaround times for business requests due to other priorities and resource constraints.
  • Lack of Business Acumen: IT may not have the capacity or knowledge to provide deep insights into specific business focus areas. This prevents full understanding of the business context and leads to potentially less relevant insights.

Because of this, the business may want the BI team to report to the business org. This is to have the team more aligned with business goals and provide more relevant and actionable insights. By becoming domain experts, this business-centric BI team can deliver quicker responses. This allows them to adapt faster to changing business needs and priorities, leading to more accurate and useful analyses.

The following illustration shows how BI would fall under the business org:

BI with Business - org structure showing Bi as part of Business

A business-centric BI Team sounds like it will solve all the challenges the business faced with BI centralized within IT the organization. However, a few potential pitfalls could exist. The decentralized control can lead to inconsistencies in data governance and security. Business teams may lack the technical expertise to build and manage complex BI tools and infrastructure.

Modern Data and Analytics Platform

Now that we have a better understanding of the common BI and Analytics roles and responsibilities, let us dive into the following illustration depicting a high-level modern Data and Analytics platform.

Data Analytics Platform - full fabric data platform

This model is based on the Microsoft Fabric platform where IT manages and owns the Data Engineering and Data Warehouse CoEs.

  • Data engineers build and manage the multi-environment lakehouses and warehouses using ELT pipelines and CI/CD for DevOps.
  • Data warehouse or BI developers build and manage the Enterprise Data warehouse.
  • BI developers, specifically within the context of Power BI, typically build and manage semantic models, reports, dataflows, etc. They serve as the BI CoE, providing reports to downstream business and analytics consumers. They are usually most successful when they report to the IT org.
  • Advanced Analytics (AI/ML) consumers build models and experiments in their workspaces, along with Analysts who use Python and SQL notebooks to explore the data and models, to provide ad hoc insights or structured reports and dashboards. These departmental and self-service Business Analytics teams usually report to the business org.

Centralized vs. De-Centralized Business Intelligence Team

This section will cover more workflow details for how Centralized Enterprise BI vs De-Centralized Departmental BI solutions and artifacts are published, managed, and consumed. The illustration below shows how a centralized IT-owned BI CoE team builds and publishes enterprise-grade models, reports, and more to various departmental Power BI Fabric workspaces. They are typically published in Power BI Apps and are read-only to the rest of the org. Downstream Business consumers and Analytics teams can consume the published read-only artifacts as a managed self-service. This approach provides great governance and standardization from the central BI CoE but comes at the cost of speed and potentially depth of insight and detailed analytics.

Enterprise BI Workflow - shows how Bi CoE manages centralized report deployment

On the other hand, when BI is decentralized as part of a Business Analytics Division, each of these new teams would be responsible for their focus areas. While this seems like a great way to provide deep domain expertise to the business and speed of delivery, the pitfall is that the BI CoE has completely been dissolved with this approach, and each siloed team is left to apply its own standards, governance, integrity, and administration. However, this can result in issues, especially if no distinction has been made regarding the Analyst vs. Developer roles.

Departmental BI Workflow - shows how Analytics teams manages de-centralized report deployment

The Relevance of the BI Center of Excellence

Regardless of whether IT or the business owns BI, its relevance as a CoE must not be forgotten. Oftentimes, as part of re-organizations, BI might be moved from IT to the business and consolidated within other Business Analytics units. As we discussed previously, these Analytics teams have a different skillset from the Enterprise BI team. So, if this move from centralized to decentralized departmental BI now requires Analysts to manage BI reporting, then the Enterprise BI CoE must continue to exist to enforce standards, governance, and overall reporting platform and infrastructure administration.

BI COE Part of Business Overseeing Business Divisions

The illustration below shows how this team can co-exist within a business org with strong ties to other Analytics divisions.

BI CoE in Business- org structure showing how BI CoE can coexist in business org with analytics teams

Whether reporting to the business or IT, this Enterprise BI CoE team is critical to success. Regardless of where this team sits, their day-to-day functions should remain the same and they should have strong ties to both the business and IT.

BiCoE_BusinessandIT shows how BI CoE is relevant to both IT and Business and must have ties to both

Enterprise BI + Business Analytics

The following illustration depicts how the Business Analytics team and the BI CoE can co-exist and work together. By adhering to a DevOps Continuous Integration (CI) and Continuous Deployment (CD) framework, the Business Analytics teams can be empowered to self-manage and create deep departmental insights and analytics by creating their domain-focused artifacts in their development workspace and leveraging source control and deployment pipelines to have their solutions gated and approved by the BI CoE in the QA workspace.

Once the BI CoE approves adherence to the best practices and standards, the artifacts can be deployed to production workspaces where they will be available for business consumers. This ensures that a technical team provides governance, security, compliance, standardization, and general report administration in a quick and prioritized manner through iterative CI/CD practices. It also gives more control of reporting to the business analytics teams that can serve as domain experts by focusing more deeply on their assigned departments.

Enterprise BI + Analytics - shows CICD workflow for how BI COE can manage iterative reviews, administration and deployment of Analytics team work artifacts.

Next Steps

Related Posts

Leave a Reply

Your email address will not be published. Required fields are marked *