If you are preparing to start a business intelligence and/or data warehousing project, then you probably already know that BI project failure is not uncommon. While there are many reasons why projects fail, we can boil it all down to a handful of mistakes that can be avoided 100% of the time. When picking your BI team demand that they exhibit the following 3 traits.
Agile Methodology
Probably the most important project planning decision to be made is the delivery methodology. The only reason for building a BI/DW system is to provide decision makers with information that will improve decision outcomes. Often, making a decision means subsequent actions will result in change. This change is ongoing and leads to more decision making. A BI system must also be capable of adjusting to change. Waterfall delivery will often not be successful for this simple reason. Taking an agile approach enables incremental collection of requirements and delivery of production ready assets without a lot of costly rework.
Experienced Implementers
It seems like anyone who has ever built a report or created a database table has BI/DW listed on their resume. Unfortunately, adding a skill to a resume doesn’t directly lead to project success. Experience, on the other hand, is critical to success. Here are a few things to consider when picking your team:
– Look for people that have a record of successfully implementing, but also have seen failure.
– When working with consulting firms, don’t accept just any resource. A strong consulting company is important, but individual consultants will make or break your project.
– Steer clear of people that are experts on every vendor platform (not humanly possible). Instead find people that are working within your target platform on a daily basis.
– Don’t confuse BI/DW with OLTP & application development. These are completely different skills.
Proper Tooling
People often get emotionally attached to their favorite tools. This is understandable, but for a BI/DW project to succeed you will need tools that are more than your chosen implementer’s favorite. Relational database, ETL, OLAP, and front-end applications all play an important role. Look for these characteristics in the chosen toolset:
– Metadata driven; agile delivery is all about flexibility. It is very hard to be flexible when every change requires a mountain of rework.
– Vendor focus is okay, but don’t put the blinders on. If you are getting advice that always seems to lead toward purchasing a single vendor’s tools, then be wary. No vendor has a monopoly on good tools.
– Available Resources; a tool is only as good as the person using it. Don’t get stuck depending on a handful of people to keep your BI/DW project moving long-term.
– Avoid the General Store; look for tools that are built for your specific needs. Purchasing features that you will never need doesn’t do much good, and often adds complexity.
– Don’t be afraid of open source. There are some great solutions available.
The right people with the right tools and the right approach will be successful. Getting started with these three core principals in place is critical, but sadly this is not always the case. If you have already started your BI/DW project and you feel like it’s on the road to catastrophe, then stop, reevaluate, and make needed adjustments today. Bring in an independent expert to help evaluate if needed. Business Intelligence is too important to your organization’s success to allow it to fail.
