Nine Steps to Successful Customer-Centric Databases

Published in April 2000

At the center of every business is the customer. Providing customers with goods and/or services is the essence of why enterprises exist. Enterprises have customer-centric data in two places:
operational and analytical (decision support) environments. Operational data is used to serve day-to-day needs and includes every activity that touches the customer. Analytical data supports the
information requirements of management and is used to monitor tactics and strategies.

This article presents a proven nine-step methodology for constructing customer centric databases:

  • Business justification
  • Identification of a champion
  • Formation of an interdisciplinary cross-functional team
  • Development of a conceptual/business data model
  • Creation of a logical model to support the process model
  • Pilot
  • Validation of the savings
  • Building the production system
  • Implementation of the production system.

Business justification

To ensure that all customer-centric projects support the organization’s objectives, the project must have quantifiable business benefit. Projected savings, increased revenue and resultant
profits should be calculated in the justification phase. Funding is then available to finance the initial project as well as subsequent projects. Table 1 is an illustration of this type of

The benefit statement would then include lost profits and the cost to replace lost households. The result is an immediate source of funds for development. Typically, the business value of the first
project exceeds the cost of the first two identified projects. This approach makes projects self-funding. The planning phase develops benefits, level of effort and business strategy. Prioritization
of projects on the basis of business benefit and urgency are outlined. After the first project is identified, the succession of projects that follow is then ordered on business need.

Identification of a champion

Visionary, high-level, influential leadership is essential for projects involving customer information. Customer information is an enterprise resource that does no belong to one owner, it belongs
to everyone. The leader must be of sufficient stature that support can be mobilized across all business units.

Formation of an interdisciplinary cross functional team

It is essential that there is a team that represents all business areas as everyone is a stakeholder. Without involvement, it is impossible to implement a customer system that satisfies
everyone’s needs. To maintain enterprise focus, an oversight committee containing executive representation from all areas is also formed. This committee manages both authority and

Development of a conceptual/business data model

Three levels of data models are used in data-driven methodologies: a high-level conceptual business model, a mid-level logical model, and a detailed physical model showing how the data will reside
in the database. These representations become the “road maps” used for all customer-related application development.

The conceptual model describes the categories of information and their inter-relationships. This top-down approach to customer information requirements provides a critical view of the business
across the enterprise in a form that makes it impossible to IT to create enterprise databases and application software. The model and supporting documentation categorizes information and describes
how the categories relate. The customer information process is documented and enables:

  • Development of business requirements for this and future projects
  • Elimination of overlapping efforts.
  • Confidence that all customer-information-related systems have common definitions.
  • Multiple simultaneous projects to occur that adhere to global standards.
  • Construction of fewer gateways and interfaces between systems.
  • Agreement on enterprise strategies.
  • Identification of perceived data integrity issues that may adversely impact reliability.

Creation of a logical model

The logical model adds specific attributes on information categories and removes duplication. Using the conceptual business data model as the foundation, the logical model is the first step in
designing a relational database. Relationship between information are clarified and expanded.

Creating a complete logical model is a lengthy exercise. The plan developed in the justification phase can be used to divide the task into manageable pieces. Depending upon the plan, the logical
model may be completed in incremental steps aligned with early deliverables, or completed in conjunction with parallel projects. The goal is to gain return on investment incrementally and have
rapid development.


Once the logical model is completed, it is necessary to prepare for implementation in the chosen technical environment. The outcome of this exercise is a physical model that describes exactly how
the database will be created with defined structures, rules and relationships that reflect the production first-cut database. From this model the required code is developed to create the database
for the pilot.

The creation of a total CRM solution is expensive, risky and resource-intensive. The pilot is a way to manage risk and develop a project plan from proof of concept that validates concepts,
approaches, benefits, performance and business value before proceeding to a production system. Adjustments can then be made. The result is lower risk and an increased probability of success.

Validation of the savings

It is important to compare the pilot results to the forecasted benefits. If, for example, Table 1 was the benefit statement, a comparison would be developed comparing the pilot to existing methods.
It is important that the project be held accountable to projected benefits. Many companies incorrectly move on without this step.

Building the production system

Results from the pilot are used to tune the approach and build a production system. The process is much like that employed in engineering where a mock-up is built and used to prove a design can be
manufactured. Using this approach, the resultant system will have industrial strength, stand up to volume tests, be timely, responsive and derive business benefit.

Implementation of the production system

Customer systems have the greatest impact to an organization. They require changes in process flows and corporate culture. The planning must be rigorous with an eye towards gaining the anticipated
improvements without disrupting existing business processes. Strategies such as making the new transparent to the old, application or division roll out, or geographic roll out need to be
considered. Training is a big issue. It is impossible to change everything overnight. Incremental success is preferable to big bang failure.

In summary

Typically, the first project is just the beginning. The savings generated in the first project should carry the cost of the first and second project. Customer-centric initiatives are replete with
opportunities for increased revenue and reduced expenses. The key is to identify the opportunities in the initial planning exercise. The project team must then be held accountable to achieve
identified results. Taking each project through the nine step process will make successive projects go faster with high payoff. This management tool provides discipline to IT projects and makes for
projects which generate cumulative annual savings.


submit to reddit