top of page
Writer's pictureKevin O'Connor

How to Manage a Successful Data Warehouse Project

As you are getting ready to begin a new data warehouse project, you probably have already spent quite a bit of time doing initial planning to define high-level use cases, get buy-in from the business, and obtain funding and allocated IT resources. Now you are ready to officially launch the project and get team working to deliver these reports and dashboards that you've been talking about for months.


So where do you start?


Here are my recommendations for managing a successful data warehouse project.

Kick-Off Meeting & Covering Objectives

I suggest conducting a formal kick-off meeting with all the stakeholders including the business, IT, and any third-party vendors or contractors. This meeting is very important to get the project started off on the right foot, establish certain ground rules and communicate to everyone the key business objectives, scope, approach, deliverables and timing of the project.


After some opening remarks and introductions, start out by covering the key goals and objectives of the project. This may seem obvious, but it's important for everyone to understand what the key business drivers are and how this solution will add value and help the organization. If possible, your most senior business stakeholder should be in this meeting to communicate this message, as it will have a much more profound, long-lasting effect on the team when it comes directly from executive management rather than an IT project manager.


Defining Roles and Responsibilities

Review all the key team members directly involved in the project and their main role. You want to make sure each team member is aware of their responsibilities. It's also helpful that everyone understands what everyone else is doing to avoid overlap and foster collaboration.


Also, get commitment from the business in advance, as they will play a critical role and be involved throughout the project. Oftentimes, business leaders outside of IT may have a perception that data warehouse projects are just for the technical team. You want to make sure they understand their role and are ready to support when needed.


Finally, highlight who the key decision makers are and define the escalation process, as you most likely will hit a few roadblocks along the way that will need to be resolved quickly.


Project Scope & Key Deliverables

Covering the project scope is critical because typically this is not communicated very effectively, and team members and the business will start to make assumption about what will be included as part of the project. Be sure to include all the data sources and ingestion frequency, any data cleansing and integration requirements, how users will access the data, data security and other key deliverables such as reports, dashboards, training and documentation.


I also like to include a one-page high-level solution architecture diagram, like the one below. This doesn't need to be complicated, but a picture can help the team visualize the solution and be used to help facilitate meetings and discussions throughout the project.

Risks, Dependencies and Assumptions

Finally, you will want to cover any potential risks and assumptions. Risk is inherent on any project and identifying these early on in a project will help set expectations and bring awareness to understand the type of problems that could delay or possibly derail the entire project. Each risk should be assigned an owner and have action items on how you plan to mitigate the risk. Assumptions and key dependencies should also be reviewed, as they might also have a negative impact to the project.


Delivery Approach & Timeline

When implementing Data Warehouses, I recommend that organizations take an iterative approach and not try to boil the ocean. You want to deliver some value to the business as quickly as possible. I suggest delivering the first set of reports/dashboards within three months or sooner. Once users can see and interact with the data, they will see value, become excited and start evangelizing the data warehouse to other stakeholders in the organization and have them yearning for more. This also helps mitigate the chances of delivering a solution that wasn't exactly what they wanted, without having to do much rework. Of course, a flexible, scalable solution will be required as more data and reports are included in subsequent releases. In short - deliver often, build on momentum and adapt along the way as needed.


Conclusion

If you plan carefully and take the time to address these aspects of your data warehouse project, you will be off to a great start to ensuring a successful data warehouse implementation.


header.all-comments


View Our Recent Posts

bottom of page