Before I became a consultant I was a classical guitarist. When I started playing, I quickly realized that I had a gift. I set out to find music that I wanted to play and skipped the rudimentary stuff that I felt was boring. That was a huge mistake. For a while I was exceptional. It wasn’t until years later when I got to the truly difficult pieces that my mistake hit me in the face. Having skimmed over the basics, when I had gotten to a professional level of playing everything took much more work. Eventually I had to stop playing the pieces I wanted to play and restart all the rudimentary stuff I had skipped before. It was a pride swallowing siege that put the e-brake on my progress. I had been so in love with the next step that I really forgot to pay attention to the step I was supposed to be on.
Many BI implementations fall into the same trap that I did. Here are some tips I give clients on how to avoid doing so.
Success is nice, but be realistic about your capabilities.
A lot of BI implementations start off simply with a star or two maybe even a cube, and that project will go well. Seeing that success they feel they can take on the world and begin undertaking immense projects thinking that there’s nothing they can’t take on.
Just because you can play Green Day like no one’s business doesn’t mean you’ll do great at Bach. Keep a close track of what your team is good at and what they haven’t done yet. Realize that just like every piece of music requires different skills so too every project demands a different set of skills, skills that might not be honed or even tested. Plan accordingly.
Be wary of fads/Live in the present
I’ll admit I’m part of this problem. Bloggers like to talk about the next big thing. Companies like to push the next big thing. And everyone wants to be in on the next big thing. Many people see the cool new technology and end up saying, “I must have that right now!” Many make the mistake of undertaking these projects without letting their prior projects properly “gel”.
It’s like building a house of cards. Groups hurry though one project to get to the next, which then gets rushed to get to the next project. Then a few months or years later someone wants to add some basic capability to the system and the developers reply that it would take a rearchitect to implement. Give your projects enough time to be properly thought through, implemented, and supported, before moving onto the next thing.
Many take this to mean that I’m saying companies should never be “early adopters”. That is certainly not the truth. Like I said in the first bit of advice, know your capabilities. If you honestly feel you have the team that can handle the challenges of an early adoption and do it right, there is absolutely no reason not to do so. Just because some pieces of music are difficult doesn’t mean that no one should play them, same thing here.
If BI is important to you, your organization should reflect that
Perhaps it’s because business intelligence has a perception as an “add-on” or “bolt-on” to transactional systems, but many IT organizations do not have someone to oversee the entirety of their business intelligence. Some companies simply let whoever needs a BI initiative deliver it. This leads to confusion and delay and is not a useful engine. Teams will build according to their project, overlooking other projects done by other teams and upcoming projects by yet other teams. It is unsurprising that these implementations turn into nightmares to support.
Not every organization can afford, or even needs, a director of business intelligence. But at the very least the work should be funneled through someone that can properly set strategy, direction and tactics for the overall BI implementation. Obviously the larger your implementation the more attention needs to be paid to it.
