Small and medium-sized businesses (SMBs) enjoy a growing marketplace for packaged business intelligence (BI) systems based on SQL Server. From all-in-one systems that you host on your own server to Software as a Service (SaaS) offerings you just sign up for, the options grow every day. The lessons huge enterprises have learned from BI over the past decade are paying dividends for SMBs, as vendors leverage those lessons to create smaller, less-expensive and easier-to-deploy SQL Server BI systems.
However, the advantage that giant enterprises have is that they’re building those systems from scratch. They don’t need to worry about comparing features and vendors, because they’re paying to develop the whole darn thing. When an SMB goes shopping for a BI system, being a smart comparison shopper is mandatory. With that in mind, here are the five key things you’ll need to pay attention to:
- Vendor experience. When looking at a BI system, see how experienced the vendor is with BI in general. A vendor that’s already making a living custom-building BI installations for huge companies is going to have the experience necessary to create SMB-focused systems. If a vendor has never done BI, ask to be shown its bona fides.
- Performance. Look for a vendor that’s willing to recommend any hardware that you’ll need to host its system, so that you’re running on a “supported” platform and can go to the vendor for help if your SQL Server BI system doesn’t deliver the performance you need. Otherwise, you’ll wind up having the hardware vendor and BI vendor pointing fingers at each other. Also, ask the BI vendor to show you some existing production systems that are sized about the same as your proposed system, and ask to be shown how those systems perform. Make that vendor prove it can handle your data.
- Output. The whole point of a BI system is output: dashboards, scorecards, reports and so on. Make sure you’ve got a robust built-in set of output mechanisms. It’s great when vendors offer “custom reporting,” but you should be able to start using a system effectively without having to immediately build a bunch of custom reports.
- Connectivity. A BI installation has to pull business data from your existing systems, so make sure you know what systems that entails, and that a proposed BI system can connect to them all. If the BI system can’t get to your data, it’s useless.
- Mechanism. There are two primary ways of turning your business data into business intelligence: a disk-based data warehouse, which operates on less-than-real-time data, and in-memory analytics, which relies more on up-to-the-minute data. Both have value and a place in your business, and ideal BI systems incorporate both. Make sure you know what you’re buying, and ask the vendor to explain how to turn data into facts.