Saturday, July 6, 2013

Business Intelligence Agile Implementation

It has been a while since my last post about the Business Intelligence. Today's post is about agile BI implementation. This article come from Tibco (My Competitor) that give me some insight and knowledge. If you are interested with the agile implementation, maybe this post will give you some hint. :)

6 Steps To an Agile Business Intelligence Implementation

Interest in using the Agile Methodology for business intelligence development and implementation is increasing. Agile methods emphasize a strong partnership between end-users and developers that can increase user adoption and help limit scope to the most critical functionality. It also helps users benefit from BI solutions faster, in smaller pieces, rather than all at once when a single solution is delivered. Ken Collier, a senior consulting with Cutter Consortium, describes six steps to achieving a successful agile business intelligence implementation.

1. Involve Users Early and Often
End users should be involved in the process from the very beginning and should stay engaged on a daily or weekly basis. This will help ensure that end-users get the functionality they need the way they need it.

2. Features First, Not Data First
Some BI implementations may begin with collecting, merging and cleaning the different data sources available within an organization. However, with Agile development, the features and functionality required by users should be defined first, followed by a strategy of how to gather the right data to deliver that functionality.

3. Prioritize Based On Value
When working with end-users to identify the functionality they need, the Agile methodology also requires you to divide that functionality into smaller deliverables over short periods of time. This encourages end-users to prioritize the features they really need that will be delivered first.

4. Test Automation
Using automated testing tools is critical to being able to deliver functionality in short iterations. Without these tools, manual testing will consume too much time in the process.

5. Encourage Collaboration Among Developers
Collaboration and cooperation among developers with different roles in a project, such as data modelers, database administrators and BI developers, are paramount. The greatest success will result when everyone on the development team works together as a single unit, regardless of their individual roles.

6. Start Quickly Out of the Gate
Everybody on the team, including end-users and developers, need to start working on their portions of the project immediately. While some of the pressure of fast deliverables can be intense, it makes for reduced stress in the long-run.