This data model represents events, entities and objects in the real world that are of interest to the company. It is subject oriented and includes all aspects of the real world, primarily activities pertaining to the business.

To use lay terms, a data model can be considered a road map to get one employee from point A to point B in the least mileage, most scenery and shortest time of travel.

In the science of computing, data models are structured and organized data structures that are implemented in a database management system. Aside from defining and organizing business data, data modeling also includes implicitly and explicitly imposing constraints and limitations on the data within the data structure.

A data model may be instance of a conceptual schema, logical schema and physical schema.

A conceptual schema is a description of the semantics of an organization. All the terms of the business from the most minute details such as staff information to the most complex business transactions are being defined and translated as entity classes. Relationships among entities are also defined in a conceptual schema.

A logical schema is a description of the semantics in the conceptual schema. It can be represented by a particular technology for data manipulation. This schema is composed of particular descriptions of columns, tables, XML tags, object oriented classes and others. Later on, these descriptions will be used in the software applications implementation to simulate real life scenario of activities in the business.

The physical schema, as the name implies, is the description of the physical means for storing data. This can include definitions for storage requirements in hard terms like computers, central processing units, network cables, routers and others.

Data Architects and Business Analysts usually work hand in hand to make an efficient data model for an organization. To come up with a good Common Data Model output, they need to be guided by the following:

1. They have to be sure about database concepts like cardinality, normalization and optionality;
2. The have to have in depth knowledge of the actual rules of business and its requirements;
3. They should be more interested in the final resulting database than the data model.

A data model describes the structure of the database within a business and in effect, the underlying structure of the business as well. It can be thought of as a grammar for an artificial intelligence in business or any other undertaking.

In the real world, the kinds of things are represented as entities in the data model. This entities are can hold any information or attribute as well as relationships. Irrespective of how data is represented in the computer system, the data model describes the company data.

It is always advised to have a good conceptual data model to describe the semantics of a given subject area. A conceptual data model is a collection of assertions pertaining to the nature of information used by the company. Entities should be named with natural language instead of a technical term. Relationships which are properly named also form concrete assertions about the subject.

In large data warehouses, it is imperative that a Common Data Model must be consistent and stable. Since companies may have several databases around the world feeding data to a central warehouse, a Common Data Model takes a lot of load in the central processing of the warehouse because disparities among database sources are already made seamless.