Case Study

Better MQL Quality: Fixing Lead Scoring and Qualification in Pardot

Lead Management
Marketing Automation
Pardot
Salesforce CRM
Professional Services
Score + Grade
MQL qualification based on both engagement and ICP fit.
1
Centralised scoring framework replacing manual scoring across assets and automations.
D → A+
New grading model introduced to measure fit against the Ideal Customer Profile.

A global research and professional services company with 2,500 employees needed a better way to qualify leads before they reached Sales. By redesigning scoring, introducing grading and improving qualification rules, the team increased trust in MQL quality and made lead prioritisation more accurate.

Challenge

Sales teams were receiving leads that looked qualified but often lacked real buying intent.

The existing scoring model gave too many points for low-value activities and reduced scores too slowly when prospects stopped engaging. As a result, inactive leads could still appear highly qualified.

The setup was also difficult to maintain. Scoring values were spread across Completion Actions, Automation Rules and Engagement Studio programmes. Even a simple scoring change required updates in multiple places.

During the review, we found another issue. Some prospects had extremely high positive scores, whilst others had very large negative scores from previous attempts to fix the model. There was also no grading system to confirm whether a prospect matched the organisation's ideal customer profile.

Solution

The goal was to make lead qualification reflect real buyer interest, not just activity volume.

  1. Centralised Scoring Framework
    A custom scoring model was built using tags, automation rules and completion actions. Marketers now apply tags, whilst the scoring logic runs centrally. This makes future updates much easier.
  1. Scoring Recalibration
    Email opens were removed from scoring and click scoring was simplified. More weight was given to activities that show stronger buying intent, such as content downloads, form submissions and lower-funnel engagement.  
  1. Grading Implementation
    A grading model was introduced to measure how closely each prospect matched the Ideal Customer Profile. Job titles were restructured into more useful qualification fields and additional profile criteria were included in the model.
  1. New MQL Rules
    The MQL trigger was redesigned around both score and grade thresholds. Leads now need to be engaged and match the target profile before reaching Sales.
  1. Score Cleanup and Stabilisation
    Historical score inflation was corrected, extreme negative scores were reduced and new controls were added to prevent future score distortion.

The team focused on fixing the scoring architecture rather than changing a few scoring values. This created a qualification framework that can be maintained and improved over time.

Results

The biggest improvement was confidence in lead qualification. Marketing gained a clearer way to measure engagement, whilst Sales received leads that better reflected both interest and customer fit. Day-to-day discussions moved away from lead volume and towards lead quality.

The scoring model stopped measuring activity. It started measuring intent.

Before the project, high scores often came from actions that had little connection to real buying interest. The system rewarded volume of activity more than qualification quality.  

After recalibration, scoring focused on behaviours that better indicate purchase intent. This reduced the impact of inflated engagement signals and improved lead prioritisation.

The grading model added another layer of quality control. Prospects no longer qualified purely because they were active. They also needed to match the organisation's target customer profile.

The centralised scoring framework also reduced maintenance work. Future scoring changes can now be managed from one place instead of updating multiple assets and automation programmes.

Sales received higher-quality MQLs because qualification now combines buyer engagement with customer fit.

Improved
MQL quality recognised by Sales during implementation.
Centralised
Scoring logic managed through a single framework instead of multiple assets
Score + Grade
Qualification based on engagement and ICP fit.

Expert Take

Scoring becomes unreliable when every click carries the same value. The real challenge is deciding which actions actually signal buying intent and building the model around those behaviours.

What if marketing ops
didn't feel like chaos?

Latest Insights

Get closer to the ideas, tools, and tactics that move the needle.

Eloqua wait steps
read more >
Good practices: Why sharing the same sub-domains for traffic and bounceback is not a good idea.
read more >
How AI Can Help Enhance Your Lead Scoring Model
read more >
Arrow rightLet’s talk about your challanges Arrow rightLet’s talk about your challanges Arrow rightLet’s talk about your challanges Arrow rightLet’s talk about your challanges Arrow rightLet’s talk about your challanges Arrow rightLet’s talk about your challanges Arrow rightLet’s talk about your challanges Arrow rightLet’s talk about your challanges Arrow rightLet’s talk about your challanges Arrow rightLet’s talk about your challanges Arrow rightLet’s talk about your challanges Arrow rightLet’s talk about your challanges Arrow rightLet’s talk about your challanges Arrow rightLet’s talk about your challanges Arrow right