Earlier this week, I started work for a new client. Without going into revealing detail, they want a decision engine.
They hired a firm to do the work, and the firm piddled around for two months, charged them thousands of dollars, and got nowhere.
The reason they got nowhere with it is that instead of trying to understand what they were building, they looked solely at the current process. Said another way, if the client wanted a trail cut through the woods, rather than build a device that clears trails and can be reused again and again, the firm went to work chopping trees and clearing brush. And got five feet into the thicket.
A decision engine, properly constructed, is really several things at once. It can be used to poll people, where each answer leads to another logical question, and the path can diverge based upon the answers you give. It can be used to help route people to the right solution, based on a series of narrowing questions. It can be used to quiz people and test their knowledge. But to see all of that, you have to climb on top and ask the question: what is this, really?
The beauty of a decision engine is that it shouldn't follow a rigid path hard-wired into the system, but should let the client re-imagine and re-tool, as needed. It should allow them to manage and map and manipulate the myriad paths to arrive at what is best for their customers, who will eventually use this system.
I'm having a great time with it. I love a good challenge!