At the beginning of a data modeling course, I find it very useful to learn more about the participants. Most people begin by mentioning their job title followed by their job responsibilities. A title implies certain responsibilities. The data analyst captures requirements, the data modeler represents the requirements, the database administrator (DBA) and developer implement the requirements and so on. However, I have noticed a trend lately where the title "data architect" is often tied to similar responsibilities as those of the analyst and modeler.
Clarify the role of the data architect. In your organization, is the data architect responsible for both data analysis and data modeling activities, or is the data architect accountable for an entirely different set of responsibilities?
What I found fascinating about this challenge was the variety of the responses. With more than 70 responses, 49 percent felt that Data Architect = Analyst + Modeler, and 51 percent felt that data architect meant something else. There are five categories:
Analyst + Modeler. A data architect is the role where one performs the data analysis activities to capture data requirements clearly, completely and correctly, and represent them in a formal and visual way through the data models. As Ben Ettlinger, lead data administrator, summarizes, "Data architect is to data analyst and data modeler as software engineer is to programmer. People used to be programmers; now they are software engineers. These are classy terms that bring a greater level of respectability to the jobs that we do."
Analyst + Modeler + More. In addition to modeling and analysis activities, the data architect has other data responsibilities. Barry McConnell, data architect, neatly reviews, "I am responsible for being the champion and protector of data from definition through interpretation."
Analyst + Modeler + Implementer. The role needs to carry through the whole software lifecycle. Meg Reichel, data architect, says, "We also create the physical design of the data models and generate all database objects to the database. Overall, we are responsible for the data from requirements to design to implementation to maintenance." Forrest Carter, data architect, says in addition to analyst and modeler, "I also assist in the ETL development efforts, making sure that the business rules are implemented correctly and clarifying issues as they arise. I often even participate in the testing efforts."
Analyst + Modeler + "Big Picture." The role performs analysis and modeling with a broad enterprise perspective. The data architect needs to be forward-thinking to encourage the business to manage data as an asset. Bharathi Sampangi, solutions architect, believes, "A data architect needs to drive not only data modeling but also needs to make sure that data integration is based on a common metadata framework and that the integrated data is presented to the business as valid information." Other terms gleaned across responses include "holistic," "future," "politician," "long-term vision," and "enterprise information model."
Something other than Analyst + Modeler. Carol Lehn, data engineer, summarizes this role, "In our organization, the term data architect implies not only a set of skills, but also the scope of responsibilities. Our data architect role is a senior-level position requiring an in-depth understanding of our business and the ability to look across the entire organization to develop and promote a strategic vision of our data environment. The role is not involved in day-to-day, hands-on project work but will provide guidance on decisions that could affect future direction, such as where or how to implement a particular application and its supporting tables. The data architect will evaluate tools and make recommendations. Below the data architect are the data and system engineers. The system engineer is a senior DBA-type of role. The data engineer owns the enterprise logical model, defines standards, drives consolidation of redundant data, enforces the strategic vision through data models and performs data analyst and data modeling roles for highly complex or critical projects."
Two main factors impact the responsibilities of the architect:
Multiple roles. William Delaney, a data warehouse/business intelligence analyst, says, "Many organizations don't see the value in a completely staffed BI team, so usually a small number of people have some generic title that doesn't always speak to their skill sets and responsibilities." Another Challenger agrees: "I am the one and only data person in my IT organization. The company I work for recognizes the value of the data function, but they don't necessarily give it the visibility it should have. I have the big picture of all our internal systems from an enterprise perspective and also get down to the data analysis and modeling of the nuts of bolts of each disparate database system."
Program versus project. "Program" refers to an overarching set of activities on a large initiative, such as data warehousing or data quality. "Project" is one specific implementation, such as a data mart or particular data quality cleansing effort. Organizations that implement programs assign the architect with more strategic and big-picture responsibility than project-level assignments.
If you would like to become a Design Challenger or submit a modeling challenge or solution, please add your email address at www.stevehoberman.com/designchallenge.htm and submit a description of the scenario.