Great post. But, why would you manually normalize data when you can build a program to take common bad values and change them to valid values. We do that for country and state (full name vs two or three letter iso), industries, revenue, employees, etc. It's tedious (let me restate TEDIOUS) to build those programs, but once they can cover most of your common issues, you can really start to focus on campaigns, or if you are purely in ops, optimizing lead generating campaign flows to speed conversion and close biz.
... View more
I'm a long-time (back to 2007) Marketo user and one of Eloqua's first customers. What I've found works best is NOT combining the demographic (or firmographic if you are doing account scoring) lead scores with activity scores into a single score. They are such different data types that they should be kept separate. We do lead decay on activity only., which is common.
This way I can let activity scores go as high as it needs to go, because a person can do the same thing over and over and that is okay, you just need to track it.
The demographic score does not change very often and we need a way to judge what is good and what is not in target for us. So, we keep this score on a 100 point normalized scale. 100 Is the max you could possibly score and we reset the score anytime a job title or other data point changes, and rescore the demographic. We just have to determine and reassess frequently, where the line falls for a lead to be marked MQL based on both scores. Usually, by consulting with our SDR and Sales team, we find the number for a demographic fit moves down, while the number for an activity score stays rather constant. Right now we're at anything above 65 on demographic and 100 on activity. But we give our team the ability to prospect into the lead database regardless of score, so we are covered if a scoring error happens.
That is a really orthogonal way of saying let them score up on activity! We've seen people with upwards of 2000 for an activity score, FYI. Insane.
... View more