6 tips on working with the non-data modeler

“I am supposed to use this fact table but can’t find any documentation on what it is.” I heard this eye opener over the cubicle wall. It is pretty disappointing from my data architect perspective. I put a good bit of effort into making data models readable and accessible. Metadata is published and easily browsed and searched from an intranet data dictionary.

Last week I blogged about the evolving role of data architects. This comment speaks directly to that evolving role. Data warehouse and data marts began their lives in the analytics world of data analysts. As this architecture matured and evolved, the world opened to a wider non-analytics audience. Data analysts and their analytics peers are intimately familiar with data marts, data models and metadata. That is not true of many of new customers accessing analytic structures.

This week I am sharing six tips on working in our evolving data architect role. Flexibility is a key characteristic that data architects must possess to remain relevant, successful and actively contributing to the enterprise’s bottom line. Most of these tips speak to flexibility from different perspectives.

  1. Everyone is not a data architect. The basic fact you need to acknowledge first is that the majority of people you interact with are not familiar with data models. This means that you need to be able to educate and hold their hands when working with your data models.
  2. Keep it relevant, timely, and focused. Real world data models do not resemble the classroom data model. They are complex in content, relationships, and context. We need to understand and focus on our customer’s needs to make data models understandable. My cubicle mate in the opening paragraph was told to use a new data structure with no guidance. My goal is not to make him an expert on our business intelligence architecture but to give him sufficient knowledge to use the data structure in an appropriate manner.
  3. It depends on the person. My training portfolio is littered with modeling 101 sessions and data modeling workshops. My employers felt the success of the data administration group was dependent on me educating and selling the development staff on the value and benefits of the data modeling lifecycle. That cookie cutter approach is not so relevant today. Education now takes on more of a mentoring look and feel based on the individual, the partner’s role and their immediate need.
  4. Walk in their shoes. Data architects MUST understand their customer base. We must concentrate on what their job entails and what makes them successful. It is about alignment. Data architects must align customer needs with the data artifacts they manage. Good customer alignment is a more effective method of selling data management. This alignment also provides the customer an on-the-job educational experience that is more likely to succeed.
  5. Make it reusable but customizable. There is a definite trick to making this relationship-based approach that aligns to the individual’s needs a success. It starts with inventorying your artifacts, services, and skills. Being able to answer the who, what, where, when and why of each of these is critical. They become the building blocks of your relationships with new customers. It is easy to tweak these building blocks based on the individual and situation. Being able to accurately articulate what value you bring to the table and how you can be a valued partner is a must.
  6. It can make your job easier. It is easy to resist change when our plates are full. Change does indeed take effort. As with any value proposition, you must weigh the benefits to the investment to change. Flexibility may not be your strength and may actually be your weakness. These tips are meant to guide you in making incremental changes to ultimately make your work easier.

Tom Bilcze

Print Friendly, PDF & Email

Leave a Reply

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