Nice Myk Pono. +1 for identifying when predictive lead scoring makes a difference. When you have just 100/leads a month you don’t need any tools — just call everyone or use simple heuristics.
As soon as you get >1,000 it gets too hard for a sales force to follow-up with everyone. People miss qualified leads or waste time with those who will never buy. You need software to help.
This is an instance where you can waste a ton of time building your own in-house regression tools. Teams who try it usually discover obvious correlations like “clicking on the sign-up button is highly predictive of buying” — technically true but obvious and in-actionable.
Companies like MadKudu do predictive lead scoring as an outsourced service and have already solved these first-order problems. There are other solutions but MadKudu is the only one I’ve personally worked with. (disclaimer: I’m also an investor)