2 of 2 people found this helpful
There are several good threads on this topic. Please do a search.
- Consider predictive and forget made up scores.
- you can erase all of the scores (or =0)
- you can then batch update or re-score some activities within up to 90 days. But consider if that activity is still fresh enough to matter.
- Demographic scoring is boring. Don't bother. Use filters instead for MQL.
- Turn off any MQL triggers or score triggers when you do run the batch update or score erasure.
Marketo has a great guide on scoring actually - The Definitive Guide to Lead Scoring
I redid our scores recently and then worked updates into the lead stages afterwards. Only once scoring is rebuilt should you determine what value an AQL has in terms of activity.
1 of 1 people found this helpful
The threshold problem is more about looking at the total count and distribution and deciding on the volume you want per month or per day.
I'd rather see use of predictive than the older method.
Bucket your scores. We originally did different scores for everything and it was hard to determine what was the right amount and how to compare. Now we use a school scoring system (A,B,C,D) and it's easy to say "well this isn't amazing, so not an A, but it's better than clicking on a CTA so better than a C...let's go with a B. Plus most sales people understand this system.
Treat customers and prospects differently - if prospects do a lot YAY. If customers do a lot...eh, that's normal.