Baby steps

It’s not easy selling into an emerging market, no matter how important it is. We won two projects in the last week or so. In both cases, the customer hadn’t heard of decision intelligence before talking with us. In one, they were looking for data analysis to guide a marketing investment; in the other, the question was to determine the effectiveness of various college course offerings.
Years ago, when we were just getting started, we would have probably lost both of these deals, because we would have pushed decision intelligence front and center. I’ve learned in recent years to take a gentler approach, however. Even if decision modeling would be a great place to start, even if doing so would end up with a more focused and cost-effective project, I’ve learned that budgets and plans created without decision intelligence in mind don’t shift too easily.
So there’s an “on ramp”. Clients are looking for a new dashboard, a great machine learning metric, an app that does optimization, an econometric model in Excel. The good news is that our team is pretty good at all of these things, so we build what they’re looking for, then add frosting to the cake by folding in a bit of an interactive decision model. Often they end up wanting more.
The vast majority of projects we’re finding are human-in-the-loop systems, which need a good mix of high-performance analytics and human expertise. This is a great challenge, because we must go beyond simple data visualization to decision visualization, and pull in principles of perceptual learning to drive a deeper collaborative understanding of how to make decisions, using this data, to drive maximum value.
So this next model will be a simple one: it’ll probably just end up as a few radio buttons for levers, three or four dependencies, and an outcome. I’ll do this in Excel, which means that I’ll be omitting the complex systems modeling side of things. But it’s a start.