In my last blog posting, you may recall, I laid the foundation for a successful BI strategy by describing the 3 components of Monitor, Analyze and Decide (MAD). After talking to a number of companies since then, it became clear to me that MAD may be leaving out a key group of users when it comes to establishing a fully sound BI strategy…the “power analysts”.
Most companies have these types of users. Sometimes it can be the CFO, or maybe even an IT analyst who really understands the business. Regardless, these types of users have a need that the MAD strategy components might overlook. That need is Deep Analytics. Sounds a bit dark and mysterious, right?! Well, I wish it were that glamorous, but it really isn’t.
In my experience, Deep Analytics can best be described as a process of performing analysis around a question, issue, or anomaly that may not be directly connected to the key performance indicators (KPI’s) that were used as the basis for your MAD strategy. In practical terms, it is the process of creating an analysis application for a power user specifically for them to do Deep Analytics.
This process has become clearer as more of our clients get further along with their QlikView implementations. The QlikView BI application makes it extremely easy (compared to traditional BI suites) to start and execute Deep Analytics. For example, let’s say that you go through the process of building out a great executive dashboard to manage your organization’s KPI’s. In other words you properly execute the MAD strategy. GREAT! Now, a tactical operational issue comes up and you turn to your power analyst for insight. With QlikView, the power analyst can use QlikView to very easily create their own application specifically to perform Deep Analytics in order to find the cause of the issue. In traditional BI suites, this scenario would probably include a requirements gathering session (or sessions), design time and an IT request to build new cubes. In the new QlikView world, no new cubes are required and it is plausible that the data needed for the analysis is already in QlikView. The amount of time needed to start the Deep Analytics process is probably measured in minutes or hours and certainly not in weeks or years.
The value to this process is SPEED. The ability to quickly zero in on a cause to take action is critical to the success of investments in business intelligence. You could also call it “time to value”…although that sounds like a polished marketing phrase. Either way, keep the process of Deep Analytics in mind as you build and execute your BI strategy and it will pay dividends.
Until next time…Shawn