The Situation
As Director of Marketing Operations at TetraScience, I inherited a lead scoring model that had drifted so far from how sales actually qualified pipeline that "MQL" had effectively become a meaningless label internally. Leads were hitting the MQL threshold and going nowhere. Sales stopped trusting the flag, which meant they stopped acting on it quickly, which made the model look even less predictive — a feedback loop that was quietly getting worse.
The Diagnosis
The scoring model had accumulated point values for campaigns that no longer existed, had no score decay, and hadn't been revisited since the company's ideal customer profile had shifted. It wasn't measuring current buying intent — it was measuring a version of the business that no longer existed.
The Approach
I re-architected the Marketo scoring model from the ground up: rebuilt point values against current engagement patterns, added score decay so stale activity stopped counting the same as recent intent, and realigned the MQL threshold against what was actually converting to Sales-Accepted Leads. Alongside the scoring rebuild, I launched a Marketing Center of Excellence to standardize how campaigns were built and tracked, and built executive QBR dashboards so leadership could see the model's performance directly instead of taking it on faith.
The Result
MQL-to-SAL conversion increased 40% compared with the previous 6-month period. Sales began trusting the MQL flag again because it started predicting what it was supposed to predict — which mattered as much for the team's confidence in the system as the number itself.
What This Doesn't Claim
This result reflects the change in MQL-to-SAL conversion rate compared with the prior 6-month period, tied specifically to the scoring model rebuild and MQL threshold realignment described above.