Random Notes on Building Business Intelligence Systems

I sat in a three-hour presentation on business intelligence systems and took exactly three lines of notes. These were:

- brand metrics vs. database marketing metrics. This is an interesting point: brand marketers traditionally look at things like awareness levels, while database marketers look at things like retention rates. Per my comments last week on the role of brands in customer experience management, we need to incorporate brand-style measures into our customer value models.

- deliver something useful every 90-120 days: this was in the context of how to keep up momentum on a long-term project. Good thing to keep in mind.

- success stories: also in the context of promoting new systems. The point is you need a few anecdotes to illustrate early success; these then will be repeated endlessly as justification for the project. The classic example is the “beer and diapers” analysis to a justify data warehouse—a possibly apocryphal tale about how market basket analysis showed that beer and diapers were often purchased together on weekends, and that sales increased when they were put next to each other. The notion of success stories—call them legends—is really important in building and maintaining system support.

Not the most productive three hours I’ve ever spent, but so it goes.