As a general rule in my opinion the effectiveness of a data model can be directly related to the accuracy and complexity of a project’s requirements. For example there is no need to work on very detailed data models when the details surrounding a specific data model have not been defined or even clarified. Developing data models when the clarity of project requirements is limited tends to introduce designed issues because the proper details to create an effective data model are not even known.

One way to avoid this issue is to create data models that correspond to the complexity of the existing project requirements so that when requirements are updated then new data models can be created based any new discoveries regarding requirements on a fine grain level.  This allows for data models to be composed of general entities to be created initially when a project’s requirements are very vague and then the entities are refined as new and more substantial requirements are defined or redefined. This promotes communication amongst all stakeholders within a project as they go through the process of defining and finalizing project requirements.

In addition, here are some general tips that can be applied to projects in regards to data modeling.

Initially model all data generally and slowly reactor the data model as new requirements and business constraints are applied to a project.

Ensure that data modelers have the proper tools and training they need to design a data model accurately.

Create a common location for all project documents so that everyone will be able to review a project’s data models along with any other project documentation.

All data models should follow a clear naming schema that tells readers the intended purpose for the data and how it is going to be applied within a project.