Modern data teams: Key roles, structures and challenges

As data-driven decision-making becomes a business norm, data professionals, not surprisingly, are in high demand: According to the 2025 CDO Study by IBM’s Institute for Business Value, nearly half of CDOs (chief data officers) say attracting, developing and retaining talent with advanced data skills is a top strategic challenge.¹

Advanced data skill sets, however, can encompass a spectrum of competencies, from programming to analytics to product development. That’s why organizations aren’t just hiring individual data practitioners on an ad hoc basis. They’re building modern data teams.

A data team is a group of people specializing in different tasks and data disciplines to deliver business value. The work of these data professionals includes:

• Organizing and improving the quality of unstructured and structured data
  
  
• Integrating data and eliminating data silos to unlock actionable insights
  
  
• Helping business stakeholders access those insights
  
  
• Spearheading analytics that illuminate past performance and forecast future outcomes
  
  
• Supporting better decision-making and faster artificial intelligence (AI) deployment

But effective data teams don’t materialize out of thin air. Organizations must thoughtfully consider the right roles, structures, priorities and tools necessary to position a modern data team—and by extension, the wider enterprise—for success.

While capitalizing on data to benefit the business is generally the overarching goal of an enterprise data team, the level of an organization’s data management maturity—its ability to leverage data to inform business decisions—can determine more specific objectives, according to Deloitte research.

The consulting firm found that less mature organizations focus on establishing foundational elements of data management—data governance, data strategy and data quality. More mature organizations, meanwhile, also emphasize data governance but their two top priorities are AI development and data products.²

With respect to AI in particular, the role of data teams is critical. According to IBV research, nearly three-quarters of CEOs say that proprietary data is the key to unlocking the value of generative AI. To ensure that proprietary data meets the needs of generative AI initiatives, data teams must address challenges ranging from slow response times to user data access requests to inaccurate and incomplete datasets.³

As technology changes, so do data teams—meaning that the composition of modern data teams has become a moving target. According to the IBV CDO study, more than 80% of chief data officers said they were hiring for data roles that didn’t exist last year, up from 60% in 2024.⁴

As data team roles and responsibilities evolve, key positions include:

• Data engineer
• Analytics engineer
• Data scientist
• Data analyst
• Business intelligence (BI) analyst
• Data product manager
• Data governance roles
• Chief data officer

Data engineers are software engineers who build and maintain an organization’s data infrastructure. They create systems to ingest information from data sources, convert raw data into usable core datasets, optimize data warehouse performance and deploy processes for data modeling and generation.

Data engineers rely on an array of technical skills—including coding and working with Structured Query Language (SQL)—to design and deploy algorithms, workflows and data pipelines that align with an organization’s data architecture requirements. Through such solutions, data engineers can help ensure data is fit for downstream use, including analysis, forecasting or machine learning.

A newer role, the work of an analytics engineer overlaps with that of data engineers, but with a greater focus on analytical modeling and business impacts.⁵ Analytics engineers deploy software engineering techniques to build systems that help stakeholders reliably derive meaningful insights from different data sources.

Their responsibilities can include data cleaning, transformation and testing; documenting key data processes; training others in how to use data; and collaborating with data scientists and data analysts to improve scripts and queries. In addition to their software engineering skills, analytics engineers also have experience with SQL, data warehousing and data visualization tools.

Data scientists apply a range of technical and business skills to extract industry-specific insights. They build statistical models, develop algorithms, train machine learning models and create frameworks, all in service of forecasting outcomes and supporting business strategies.

Much of their work revolves around the data science lifecycle—an iterative process that includes identifying opportunities or problems, data mining, data cleaning, feature engineering, predictive modeling and data visualization. Their technical skills include coding in programming languages such as Python, SAS, R and Scala, which enables them to write programs and algorithms that automate data processing and calculations.

Data scientists can work with big data platforms such as Apache Hadoop or Apache Spark. They deploy various tools and techniques for preparing and extracting data, ranging from databases and SQL to data integration methods.

Data analysts support data-driven decision-making, client engagements and business operations through their reporting, data mining and data visualization skills.

Data analysts’ responsibilities can overlap with those of data scientists, particularly with respect to exploratory data analysis and visualization. And many functions of data analytics—such as making predictions—depend on machine learning algorithms and models that are developed by data scientists.

However, while data scientists use advanced techniques to manipulate data, data analysts work with predefined datasets to draw important conclusions that inform business decision-making.

In their work with datasets, data analysts typically rely on four types of data analytics: descriptive analytics (providing an understanding of past performance); diagnostic analytics (identifying causes of past performance); predictive analytics (forecasting future trends and outcomes); and prescriptive analytics (providing recommendations for decision-making).

In their day-to-day work, data analysts use tools such as statistical analysis software, database management systems, BI platforms and data modeling and visualization tools.

Business intelligence analysts have skills and responsibilities that overlap with those of data analysts, but their roles are different in that—as the position’s name suggests—BI analysts focus specifically on business impacts and strategies. The work of BI analysts, also known as BI specialists, can improve business decisions by identifying problems, market trends and new opportunities for revenue generation.

BI analysts parse data from data sources such as relational databases and data warehouses to prepare market intelligence and financial reports. They apply statistical analysis methods to visualize data, create interactive dashboards and present their findings.

Data product managers oversee data products, which are reusable, self-contained packages that enable everyone at an enterprise—including non-technical users—to extract value from data. Data product managers shepherd the development and optimization of data products that meet a variety of business needs and apply to different use cases.

In doing so, data product managers help organizations overcome the complexity and data silos that often prevent business users from fully engaging in data-driven decision-making.

Researchers have found that successful data product managers possess a combination of technical knowledge, business acumen, communication skills, leadership qualities and an aptitude for problem solving.

They should understand technical processes and concepts such as data modeling, warehousing, integration and analytics, while also being knowledgeable about a business’s objectives, opportunities and challenges. They are also able to use agile and iterative methodologies to continuously improve data products.⁶

Data governance is the data management discipline that focuses on data quality, security and availability at an organization. Organizations’ increasing focus on data governance has led to the establishment of cross-functional data governance teams, including steering committees, data owners and data stewards.

• Steering committees: Committee members, which may include high-level executives, oversee data governance strategy.
  
  
• Data owners: Data owners oversee specific data domains across business units. They are responsible for maintaining data accuracy and offer input on data governance solutions.
  
  
• Data stewards: Data stewards implement rules outlined in organizations’ data governance frameworks through duties such as defining data quality metrics and managing metadata.

A chief data officer is an executive responsible for maximizing business value from enterprise data. A CDO typically sets an organization’s data strategy and oversees data management functions, including data governance, data analytics and data security.

Initially, CDOs were data leaders who concentrated primarily on data governance and compliance with regulatory mandates. In recent years, however, CDOs have pivoted their focus to driving business value, with experts advising that they establish KPIs that link data initiatives to outcomes such as increased sales conversions, reduced customer churn or operational cost savings.

According to IBV’s 2025 CDO Study, 92% of CDOs surveyed reported that their success depends on being oriented toward business outcomes and 85% said they could articulate how data priorities facilitate important business outcomes.⁷

Similar data leadership roles include “director of data” and “head of data.” CDOs generally report to other C-suite leaders, namely CEOs or chief information officers (CIOs), according to a 2025 survey by Deloitte.⁸

While the size of a data team can vary widely—one analysis found that data teams make up anywhere between 1 to 5% of the companies’ total headcounts⁹—the way they’re organized usually comes down to one of three data team structures:

Once an organization has selected which roles and structure will define its data team, it might encounter a number of other challenges.

Finding qualified data professionals is getting harder. More than three-quarters of CDOs surveyed by IBV said they were struggling to fill key data roles. Just 53% said their recruiting and retention efforts yielded the experience and skills necessary to achieve their business and data objectives—down from 75% the year before.

Hiring hardships notwithstanding, data teams are growing. The 2025 Deloitte survey found that 54% of CDOs reported their teams had expanded in the last 12 months.¹²

Overlaps in responsibilities and skills among data team members are common, posing significant data management challenges for enterprises—when it’s unclear which team member is responsible for which task, delays and productivity blockers can ensue.¹³

Companies can address this concern by establishing detailed descriptions of how each role should work within their own organization.

Ensuring data quality is a foundational element of data management, but many enterprises continue to struggle with the task. CDOs surveyed by IBV cited data accuracy, integrity, completeness and consistency as among the top barriers limiting their use of enterprise data.¹⁴

AI agents can help data teams improve data quality and provide transparency into data movement. Capabilities include autonomously cleansing data and detecting anomalies; validating data against predefined rules and standards; executing end-to-end lineage tracking throughout the data lifecycle; and synchronizing information across systems.