Introducing 10 Best Practices To Implement A BI Solution

Author: 
Jim Sun

Business Intelligence systems (BI) are quite a hot subject these days, companies of all sizes are using BI to help them better manage their data and thereby improve their business performance. But as far as I see, no BI solution is truly of the "plug and play" kind; there is always implementation work required for BI system users. And the implementation is not just about hiring several IT guys and letting them handle whatever the work is. A successful implementation requires cooperation among the software vendor(s), client company, and 3rd-party consultants (if there are any). In this blog, I would like to introduce the top 10 key best practices in a BI solution implementation, although I am more focused on SAP BI and Hadoop Big Data, I believe these best practices below can also be applied to other BI systems.

  • Understand the difference between a BI project and BI program: before starting a project, you need to ask yourself a question: is this s a project or a program? Basically, a project has 1 specific purpose and will be finished in a short time frame. Whereas a program has broad purpose and may take a very long time. Most of the time, a program can contain multiple projects. For example, to implement a BI solution could be a program, whereas it may contain a BI upgrade, migration, reporting and other projects.
  • Roadmap/Conceptual architecture: as we said, a BI environment may consist of multiple projects, so a best practice to make sure your BI program is not off the target is to design a program/project roadmap and conceptual architecture. These documents coordinate all the projects in the BI implementation program and describe the high level blue print for enterprise applications.
  • Business requirements from start to end: business requirements always go beyond the project implementation. So to ensure the success of the implementation, the CIO had better talk with other members in the management team to align the system better with your business requirements.
  • Scope: my last project is a good example: they wanted to implement the SAP Netweaver system initially so they hired some SAP Netweaver BW consultants, then they found that SAP BOBJ (BI) can also provide some needed functionality, then within the Netweaver project, they hired some other BOBJ consultants, that resulted in 1 more year delay for the project schedule. And then because of the high cost of hiring consultants, they decided to decrease the number of consultants and hire some internal employees to join the IT team. That project was designed to be done in about 2 years, but it turned out to be a huge 4-year project, and still has not gone live. So defining a clear project scope is quite important to save you time and money.
  • Data quality: actually, data quality is not quite related to BI implementation. But you still need to make sure that the source data is managed properly, and the reason is obvious; high data quality can save you a lot of time loading data and generating reports in the BI system.
  • Bigger is not better: unlike your source system, the purpose of BI solution is not about processing huge volumes of data, but generating meaningful and insightful findings from the data. So my suggestion is to optimize the calculation/aggregation engine of your BI system and align the development closely with business requirements.
  • Security: I would not say your BI system will be hacked from external attacks, but security is still an important factor to consider. It requires a lot of planning. You need to plan your needs for authentication and authorization rules and how it will be implemented in your, let's say, CMC. And there should be adiministrator handling this part of job.
  • Migration/Upgrading: normally a client needs to hire external resources to help them perform the migration of their BI solution. Migration and upgrading are not like downloading a new version of software and copying & paste everything from a previous version to the new version. You can learn much more about what this involves by clicking here to check the /blog/introducing-top-5-factors-consider-bobj-migration.
  • Training: remember to make sure your end-users understand how to use the cool BI stuff after the implementation. You could get external training resources to do this or you can also get your consultant/technical team help do the training.
  • Support: normally, you can get online support from the software vendor, but it is still a better idea to hire a system administrator and get him/her onsite, some companies even keep a BI team in the organization to help maintain the system and fix any issues.

implementing bi

Of course, there are also lots of other key factors to ensure a successful BI implementation, such as enabling self-service, change management and so on. ATCG Solutions as a BI implementation consulting firm can help you analyze your needs and do the implementation according to your requirements, if you want to learn more about BI system implementation or about us, please click the button below.

Have you ever read a blog and wanted to bookmark it but forgot? Or maybe you're like a lot of people and tend to keep dozens of windows open, only to experience a crash or otherwise lose track of what you were doing? Simple solution is to print this out, and that is why we provide a PDF version for later use, which you can get by pressing the button below.