Leap before you can land
Leap before you can land…
Research scientists at the Southern Illinois University have concluded that primitive frogs learned to jump before they learned to land. The result being a rather inelegant crash landing. I’m sure the frogs were happy that they achieved their goal of getting from A to B, but I do feel sorry for them having to endure the presumably painful landing, not to mention the indignity of having the rest of the pond witness the proceedings!
What struck me about this story, aside from the obvious frog reference (albeit not purple!), was the parallel to Business Intelligence projects that we encounter time and time again. BI systems that are put in place, but whilst the project teams are basking in the glory of accomplishing their goal, often sporting a couple of bruises picked up along the way, the wider business looks on not really knowing what has just happened or understanding how the company or their department are going to benefit from this leap in technology.
Although BI, as both a business function and technical toolset, has long since reached maturity, its adoption within SMEs is still surprisingly in its youth. Most companies who embark on a BI project seem to do so either because they have been sold on a fantastic vision of the future of information management, or because they have a specific business need to solve. Whatever the reasons for initiating a BI project, there is often very little thought given to how this new information system is likely to impact the wider business, or how the system can be utilised to provide far more benefit than the original project scope called for.
There are two aspects to this impact, the pain and the gain.
When the company’s data is cleaned, consolidated and remodelled to make reporting and analysis more simple (one of the primary goals of a BI project), the tendency is to use that system to reproduce the company’s old static reports but in a slightly more glossy fashion, only faster and more accurate. This stops far short of what a good data warehouse or cube is capable of. Enter self-service BI.
Self-service BI essentially allows your employees to have a level of (secured) access to the underlying BI data, combined with the right tools to perform their own analysis. This enables them to not just consume their old static reports, but to pro-actively go digging for information, knowledge and a greater understanding of your company. Gone are the days when requiring a change to a report has to be documented, requested, specified, programmed, tested & released. A well designed data warehouse and cube, along with the right client tools (often just Excel), enables a user to just do it themselves with the bare minimum of training (if any), whilst retaining the confidence that the resulting report is accurate; the warehouse/cube take care of ensuring the underlying business logic is right.
A data warehouse is, by definition, a combined set of data from a variety of source systems. This means that any change to the source systems must be assessed for their impact on the warehouse. After putting a sparkly new BI infrastructure in place, you don’t want it to fall over because someone changed the structure of a source database or extract file! This adds a layer of change control that may not have existed before, and can cause some headaches.
There is usually a moderately large requirement for business users’ time, both to help the BI architect to understand and model the business properly, and also in testing.
Pain but no gain?
So, if the benefits such as self service BI (with dashboards, data mining, KPIs, PowerPivot etc. etc) are not adopted and utilised, the business is left having made an impressive technology leap, but is then stuck with the impact on source systems, a bill, and confusion from staff who don’t see the benefit of getting their old reports via a different method.
Evolution taught the frog that by landing properly, he can quickly jump again to go even further, and avoid the pain in the meantime.