While taking questions after a recent presentation, I was asked whether it is better for a would-be fundraising data miner to start out with a fundraising background and learn the stats and modeling along the way, or vice-versa. I readily answered that it’s better to build on a knowledge of fundraising, rather than have all the technical know-how without knowledge of fundraising.
On reflection, I’m not so sure one path is superior to the other. But I’ll stick with what came out of my mouth that day, as the true expression of how I feel.
Think of it this way. Before we set out to build any model, we need to define the business question. This question, or problem, comes first. It will decide what we are trying to predict. The technical stuff comes later. It’s not that the technical stuff is less important or less difficult. It’s that it’s not the thing that sets the process in motion. With a clear objective in sight, we are motivated to acquire the tools and skills we’ll need along the way.
Framing the question happens outside of the computer, in the real world. In the same way, if we bring a fundraiser’s attitude to modeling, we will ask the right questions, build better models, and focus our energies in the direction of success.