How to analyze behavioral lead scoring model to identify improvements?

Highlighted
Level 2

How to analyze behavioral lead scoring model to identify improvements?

What recommendations or best practices do you have for analyzing an existing lead scoring model towards identifying ways to improve? 

Context: 

  • We're planning to launch a new marketing tactic that could increase the volume of leads and therefore volume of MQLs we pass to our sales team. We're already struggling with the quality of current MQLs; sales reps often neglect outreach/follow up because, MQLs aren't currently converting at a rate that makes it worth their time. We see a lot of value in increasing our MQL conversion rate so that our sales team prioritize outreach. 
  • We've been running a lead scoring model for about 2.5 years via an operational program we created in partnership with a consultant who helped us launch our instance. 
    • Our model was a starting point very much based on our "best guess." At this point, I'd like to believe that we've gathered enough data that we should be able to analyze and tweak the scoring model so that it does a better job of bubbling up leads that will become sales opportunities. 
  • High level, when leads reach our scoring threshold, they become MQLs and we flag them for our sales team.
  • Conversion out of MQL status occurs when MQL is added to a Salesforce opportunity.
  • Currently only running behavioral scoring. Adding in demographic scoring is on the to-do list towards improving our scoring model. 

Questions I'm interested in answering include:

  • Are we awarding too many points for any given behavioral scoring activity?
  • Are we awarding too few points for any given behavioral scoring activity?
  • If we increase our score threshold by X pts, can we predict what the impact will be on conversion rate?

Thanks for any help you can offer!