The actual task of Business Intelligence (BI) is business data analysis. The typical operation is a query on a massive database of business data, usually fed from existing operational systems like order entry or sales, manufacturing or production, and distribution or delivery. For the SMB, this means that existing systems must add ETL (extract, transform, load) links to a central database that is strictly for data analysis. This deployment can take place pretty quickly, but typically, even in the SMB case, it will take a month or more to really start using BI.
The key question here is whether you anticipate growing into a large enterprise soon or not. If not, a SaaS solution or one that is pre-packaged and easy to query with is the best option. If you do anticipate being a large enterprise (say, more than 500 employees) in the next couple of years, take the extra time and do either open-source (if you have tech-savvy IT) or global BI.
Almost all businesses, SMB or not, have a wish list of things that they would like to find out about their customers, their operations, and their suppliers, beyond what their existing reports tell them. The SMB should take this wish list, prioritize it, and be ready to hit the ground running when the BI system has enough data in it to be useable. Again, the key is to be flexible: the initial answers you get may not be what you expect, and you will need to formulate follow-on questions quickly. Once beyond the initial shock, however, BI becomes quite routine.
One final point: as with other products, there must be high-level business buy-in. Unlike other products, that means not only that you deliver routine reports but also that you deliver periodic warnings and unexpected findings, and that the high-level business executive pays attention to these.
If you are holding on to the idea that meetings have to be held in a conference room, it’s time for you to reconsider. more