.png)
Read Time: 3 Minutes
Most Account Engagement scores are decorative... they look smart in dashboards but predict nothing.
Short answer: Account Engagement scoring predicts closes when grade and score work together. Set the score on behavior, set the grade on fit, and only sync to Salesforce when both clear your threshold. Add decay rules so old activity stops inflating scores, and audit weekly with a closed-won regression report.
Sales says marketing leads do not close. Marketing says sales does not work them fast enough. Both teams are right when scoring is broken because the score that triggered the handoff did not actually predict intent.
Score on behavior, grade on fit
The score and grade do different jobs. Stop combining them.
- Score is behavior. Pageviews, form fills, webinar attendance, demo requests, pricing page visits.
- Grade is fit. Title, industry, company size, region, tech stack.
- A high score with a low grade is a curious intern reading every blog post on a free trial. Do not pass to sales.
- A high grade with a low score is a perfect-fit prospect who needs nurture, not a call. Drop them in a sequence.
- Only the high-score and high-grade leads earn a rep. Everything else is a waste of an SDR hour.
Set decay so old clicks stop lying
Behavior 90 days ago does not predict a close today. Without decay, your top scorers are anyone who downloaded an ebook 6 months ago and forgot you exist.
- Open Account Engagement Settings, then Scoring Categories.
- Add a decay rule that subtracts 5 points every 30 days of inactivity.
- Cap any single asset at 25 points so one whitepaper does not max out the score.
- Reset score to 0 if a prospect is inactive for 180 days.
- Test the decay on 50 prospects before you turn it on across the database.
Only sync to Salesforce above the threshold
Most orgs sync every prospect. That is why Salesforce is full of dead leads and your reps stop trusting marketing.
- Set a threshold like score 100 plus grade B or higher.
- Use a Completion Action to assign to a queue only when the threshold is hit.
- Anything below the threshold stays in Account Engagement nurture and never touches a rep.
- Add a Salesforce field called MQL Reason so reps see what triggered the handoff.
- Add a second field called MQL Score Snapshot so you can audit what the score was the day they came in.
Audit scoring against closed-won every month
Run a regression report every month so the scoring model stays honest. This is the step every team skips and it is the reason scoring drifts.
- Pull every closed-won opportunity from the last 90 days.
- Trace each one back to its Account Engagement score at the moment of MQL handoff.
- If most closed-won had scores under 75, your threshold is wrong and you are losing real deals to the nurture bucket.
- If high-score leads are not converting, your behavior weights are wrong. Demo requests should weigh more than blog reads.
- Adjust 1 weight per month. Never tune 5 things at once or you lose the ability to tell what worked.
Bottom line
On Monday, pull every prospect synced to Salesforce in the last 30 days. Filter to score under 50. That number is the size of your noise problem. Fix the threshold this week.
Frequently asked questions
What is the difference between Account Engagement score and grade?
Score measures behavior like pageviews, form fills, and demo requests. Grade measures fit like title, industry, and company size. Score is what they do. Grade is who they are. You need both to predict a close.
What threshold should I use to send a lead to sales?
Most B2B teams land between 75 and 150 score plus a B grade or higher. Start at 100 plus B and adjust based on a closed-won regression report after 30 days. The right threshold matches where your closed-won leads were at handoff.
How often should I update my Account Engagement scoring rules?
Audit monthly. Adjust 1 weight per month, not 5. Big changes break the trend line and you lose the ability to tell what worked. Small changes compound.
Why are my Account Engagement leads not converting in Salesforce?
Three usual suspects. The threshold is too low and you are passing fit-only leads. There is no decay so old clicks are inflating scores. Or sales has no SLA on MQLs and they sit for days. Check all three before you blame the scoring model.
Should I score every email open?
No. Email opens are noisy because of inbox preview panes and mail clients fetching images. Score clicks, not opens. A click is intent. An open is not.
Want more of this? The Simply Scale newsletter drops every week: https://www.gosimplyscale.com/newsletter
