top of page

The Pitfalls of Data Warehousing and How to Avoid Them

Depending on the needs of the enterprise, a data warehouse is a viable data storage solution that also offers a great deal of organization for your data. Because of this, there is so much complexity involved in setting one up. Without a proper understanding of the data warehouse structure, organizations that will use it and the people building this data structure will come across critical pitfalls that ultimately prevent the data warehouse from reaching its full efficiency potential.

Here are a few critical and common data warehouse pitfalls that organizations should be on the lookout for and must address immediately as soon as they come up at any time during the data warehouse setup.

Lack of input from critical personnel with experience in data warehouses

Given their complexity, data warehousing projects are fundamentally different from typical operational projects. An experienced data warehouse project leader understands these facts and keeps the vision of the project in concert with the real-world reality of decision support. In addition, the expertise of a data warehouse architect is needed in designing a scalable, robust, and maintainable architecture that can accommodate the expanding and changing decision support requirements.

It also helps to have highly experienced, senior-level individuals who can identify problem areas in the organization that can be dealt with early on to ensure the data warehouse is successfully implemented.

Lack of understanding of the concepts and tools involved in data warehousing

For a data warehouse project to succeed, users should be educated to convey the concepts of the data warehouse and other related terms and technologies.

In particular, data acquisition developers might need to be trained on a transformation tool while data warehouse access developers will require significant training in an OLAP tool, and data administration developers will need training on a tool that will integrate all of the company’s metadata into one repository. There might also be a web component used to access the data warehouse and the metadata repository that users might need to familiarize themselves with. Depending on your organization, additional training and outside consulting could be needed for each of these areas.

Keep in mind that these are only data warehousing-specific training issues. There still needs to be an understanding of the hardware, middleware, desktop, RDBMS, and coding language of the transformation tool.

Taking on an ambitious multiyear project rather than pursuing more manageable iterative development efforts

Data warehousing projects stretch an organization in ways unlike that of operational systems projects. For one, an enterprise data warehouse requires consent and commitment from all key departments within a corporation. Complicating this is a sadly common fact that the learning curve of the decision support project team is seldom understood or planned for. There will always be new tools that need to be used, thus requiring tool-specific training.

As such, adding massive amounts of data into the equation significantly increases the points of failure the organization cannot afford to experience. Data warehouses are best built iteratively, which means the highest probability for success comes from implementing a decision support system in a phased approach. The first iteration provides an opportunity to set the stage for bigger and better future implementations.

Being overly enamored with technology and data rather than focusing on the business’s requirements and goals

Some data warehousing providers tend to focus more on data architectures, methodologies, and technologies in order to create what they would consider a “perfect” data warehouse, but at the cost of ignoring what their clients actually want. As the client, it is important to emphasize your needs for the data warehouse and that they are effectively addressed even if the architecture, methodology, or technology is not as “cutting-edge” as providers want it to be.

Paying more attention to ease of development for those building the data warehouse than ease of use for the end-user

Some developers and data architects design their data warehouses in such a way that it would be easy for them to set up and complete while being oblivious to the end-user experience or the quality of the virtual servers being utilized by the virtual servers. Above everything else, a good data warehouse is one in which users can easily access the data stored there, with quick loading times and good query performance housed in reliable virtual servers. There should also be training, online tutorials, and documentation that is readily available for all users regardless of their experience with data warehouses in general.

Presuming that the business, its requirements and analytics, and the underlying data and the supporting technology are static

Some data warehouse architects and developers build data warehouses with the false assumption that the requirements and analytics being used by the business will remain constant. It is already a given fact that no business remains static. Poor database modeling often involves developing repeated data structures along with hard-coded names and inconsistent naming conventions.

While striking a balance of creating the right-size fact and dimension tables or the degree of normalization is no small feat; exercising common sense to develop data structures that are scalable enough to adapt to the evolution of the business is important.

Failing to assess the acceptance or readiness of the organization

While the benefits of the data warehouse are immense for any organization that may choose to adopt it, the success of its implementation will largely depend on user acceptance and readiness. If the users do not yet accept the core belief that the data warehouse is a foundation towards improved decision-making, the efforts in implementing a data warehouse will all just be an exercise in futility.


bottom of page