lead-scoring skill

---
name: lead-scoring
description: "Set up and automate lead scoring for HubSpot and other CRMs. Use when a user wants to score leads, define MQL/SQL criteria, build scoring matrices, configure lifecycle stages, implement engagement scoring, or automate lead qualification. Instruction-only skill with scoring frameworks and step-by-step HubSpot setup guides."
metadata:
  openclaw:
    requires:
      env:
        - HUBSPOT_ACCESS_TOKEN
    primaryCredential: HUBSPOT_ACCESS_TOKEN
    credentialNotes: "Required for HubSpot API access to configure scoring properties and workflows. For Salesforce, set SALESFORCE_ACCESS_TOKEN instead."
---

# Lead Scoring Autopilot — AI-Powered Scoring for HubSpot & CRMs

## Overview

Lead scoring is the process of assigning numerical values to leads based on their likelihood to convert into customers. This systematic approach helps sales and marketing teams prioritize their efforts on the most promising prospects, dramatically improving conversion rates and ROI.

This skill provides you with frameworks, templates, and automation tools to implement comprehensive lead scoring across major CRM platforms, with special focus on HubSpot integration.

## Table of Contents

1. [Understanding Lead Scoring](#understanding-lead-scoring)
2. [Lead Scoring Components](#lead-scoring-components)
3. [Setting Up Your Scoring Model](#setting-up-your-scoring-model)
4. [Platform-Specific Implementation](#platform-specific-implementation)
5. [Advanced Scoring Techniques](#advanced-scoring-techniques)
6. [Monitoring and Optimization](#monitoring-and-optimization)
7. [Common Pitfalls and Solutions](#common-pitfalls-and-solutions)

## Understanding Lead Scoring

### What Makes a Good Lead?

Before diving into scoring mechanisms, you need to understand what makes a lead valuable to your business. Great lead scoring combines two critical dimensions:

**Explicit Scoring (Demographic Fit)**
- Company size, industry, location
- Job title, seniority, department
- Budget indicators, technology stack

**Implicit Scoring (Behavioral Engagement)**
- Website activity, content consumption
- Email engagement, social media interaction
- Sales interaction history, meeting attendance

### Lead Scoring vs. Lead Grading

Many organizations confuse scoring with grading:

- **Lead Score**: Measures interest level (behavior-based, changes frequently)
- **Lead Grade**: Measures fit (demographic-based, relatively static)

Combine both for maximum effectiveness: A+25 means excellent fit with high interest.

## Lead Scoring Components

### 1. Demographic Scoring (Fit Score)

#### Company-Level Attributes

**Industry Scoring** (0-20 points)
- Perfect fit industries: +20 points
- Good fit industries: +10 points
- Poor fit industries: -5 points
- Exclude list industries: -50 points

Example for B2B SaaS:
- Technology/Software: +20
- Professional Services: +15
- Financial Services: +15
- Healthcare: +10
- Retail: +5
- Government: -5
- Non-profit: -10

**Company Size Scoring** (0-25 points)
- Ideal size range (e.g., 100-1000 employees): +25
- Acceptable range (50-99 or 1001-5000): +15
- Too small (<10 employees): -10
- Too large (>10,000 employees): -5

**Revenue Indicators** (0-20 points)
- Public revenue data in target range: +20
- Funding announcements (Series B+): +15
- Fast-growing company indicators: +10
- Financial distress indicators: -15

#### Individual-Level Attributes

**Job Title Scoring** (0-30 points)
- Decision makers (CEO, CTO, VP): +30
- Influencers (Director, Manager): +20
- Users (Coordinator, Specialist): +10
- Students, job seekers: -10

**Seniority Levels** (0-15 points)
- C-level: +15
- VP level: +12
- Director level: +10
- Manager level: +8
- Individual contributor: +5
- Intern/entry level: +2

**Department Relevance** (0-15 points)
- Primary buying department: +15
- Secondary influence departments: +10
- Unrelated departments: +2
- Departments that typically block: -5

### 2. Behavioral Scoring (Interest Score)

#### Website Engagement

**Page Visit Scoring** (1-10 points per visit)
- Pricing page: +10 points
- Product demo page: +8 points
- Case studies: +6 points
- Blog posts: +2 points
- Careers page: -2 points
- Multiple visits to same page: diminishing returns (50% after 3rd visit)

**Time on Site** (0-5 points)
- >5 minutes: +5 points
- 2-5 minutes: +3 points
- 30 seconds-2 minutes: +1 point
- <30 seconds: 0 points

**Download Actions** (5-20 points)
- White papers: +15 points
- Product datasheets: +12 points
- Case studies: +10 points
- Blog content: +5 points
- General resources: +3 points

#### Email Engagement

**Email Interaction Scoring**
- Email open: +2 points
- Link click: +5 points
- Multiple link clicks: +3 points each
- Forward/share: +8 points
- Reply: +15 points
- Unsubscribe: -10 points
- Marked as spam: -20 points

**Email Campaign Performance**
- Opened all emails in sequence: +10 points
- Clicked multiple campaigns: +15 points
- Progressive engagement (opening more over time): +8 points
- Declining engagement: -5 points

#### Social Media Engagement

**LinkedIn Activity** (2-10 points)
- Company page follow: +5 points
- Content share: +8 points
- Comment on posts: +10 points
- Direct connection request: +12 points

**Twitter Engagement** (1-5 points)
- Follow company account: +3 points
- Retweet content: +4 points
- Reply to posts: +5 points

#### Event Participation

**Webinar Engagement** (10-25 points)
- Registration: +10 points
- Attendance (full): +15 points
- Partial attendance: +8 points
- Q&A participation: +5 points
- No-show after registration: -3 points

**Trade Show/Conference** (15-30 points)
- Booth visit: +15 points
- Demo request: +25 points
- Literature request: +10 points
- Business card exchange: +20 points

### 3. Intent Signal Scoring

#### Third-Party Intent Data

**Research Activity** (5-20 points)
- Researching your solution category: +15 points
- Researching competitors: +10 points
- Reading comparison content: +12 points
- Looking at implementation guides: +20 points

**Technographic Changes** (10-25 points)
- Adding complementary technologies: +15 points
- Removing competing solutions: +25 points
- Technology stack expansion: +10 points
- Infrastructure investments: +12 points

#### First-Party Intent Signals

**Content Consumption Patterns** (5-15 points)
- Bottom-funnel content (ROI calculators): +15 points
- Implementation content: +12 points
- Comparison content: +10 points
- Educational content: +5 points

**Search Behavior** (3-12 points)
- Branded searches: +12 points
- Solution category searches: +8 points
- Implementation-focused searches: +10 points
- Problem-focused searches: +5 points

## Setting Up Your Scoring Model

### Phase 1: Historical Analysis

Before implementing lead scoring, analyze your existing customer data:

1. **Customer Profile Analysis**
   - Export all customers from last 12 months
   - Identify common demographic attributes
   - Note typical engagement patterns before conversion
   - Calculate average deal size by customer type

2. **Lead Source Performance**
   - Analyze conversion rates by traffic source
   - Identify highest-value lead sources
   - Weight scoring based on source quality
   - Account for lead source in initial scoring

3. **Sales Team Input**
   - Interview sales team on ideal customer profiles
   - Understand lead qualification criteria
   - Identify common objections and blockers
   - Gather feedback on lead quality by attribute

### Phase 2: Model Design

**Step 1: Define Scoring Ranges**
- Cold leads: 0-30 points
- Warm leads: 31-70 points
- Hot leads: 71-100 points
- Sales-ready leads: 100+ points

**Step 2: Assign Point Values**
Use the 100-point scale as your foundation:
- Demographic attributes: 40% of total score
- Behavioral signals: 50% of total score
- Intent signals: 10% of total score

**Step 3: Create Decay Rules**
Not all activity should count forever:
- Website visits: Decay 50% after 30 days
- Email engagement: Decay 25% after 60 days
- Content downloads: No decay for 90 days
- Event attendance: No decay for 180 days

**Step 4: Negative Scoring**
Implement negative scoring for:
- Job titles that never buy (students, interns)
- Companies outside target market
- Unsubscribe/spam activities
- Competitor employees
- Inactive engagement (no activity for 90+ days)

### Phase 3: Lifecycle Integration

**MQL (Marketing Qualified Lead) Criteria**
- Lead Score: 70+ points
- Demographic Grade: B+ or higher
- Recent activity: Within last 30 days
- Required information: Email, company, role

**SQL (Sales Qualified Lead) Criteria**
- Lead Score: 85+ points
- Demographic Grade: A- or higher
- Budget qualification: Completed
- Timeline: Within 6-12 months
- Decision-making authority: Confirmed

**Opportunity Creation Criteria**
- Lead Score: 95+ points
- All qualification criteria met
- Discovery call completed
- Budget and timeline confirmed

## Platform-Specific Implementation

### HubSpot Lead Scoring

HubSpot offers native lead scoring with custom properties and workflows. Here's how to set it up:

**Step 1: Create Scoring Properties**
1. Go to Settings > Properties > Contact Properties
2. Create custom number properties:
   - "Lead Score" (number field, 0-200 range)
   - "Demographic Score" (number field, 0-100 range)
   - "Behavioral Score" (number field, 0-100 range)
   - "Last Score Update" (date field)

**Step 2: Build Scoring Workflows**
Create separate workflows for each scoring component:

*Demographic Scoring Workflow:*
- Trigger: Contact is created or updated
- Conditions: Check company size, industry, job title
- Actions: Set property value for demographic score
- Re-enrollment: Yes (when property changes)

*Behavioral Scoring Workflow:*
- Trigger: Contact activity (page views, email opens, etc.)
- Conditions: Activity type and recency
- Actions: Increment behavioral score
- Re-enrollment: Yes

*Score Decay Workflow:*
- Trigger: Daily at 9 AM
- Conditions: Last activity date > 30 days ago
- Actions: Reduce behavioral score by 25%
- Re-enrollment: Yes

**Step 3: Create Lists**
Build smart lists based on lead scores:
- Cold Leads (0-30 points)
- Warm Leads (31-70 points)
- Hot Leads (71-100 points)
- MQLs (70+ points + recent activity)
- SQLs (85+ points + qualification)

### Salesforce Lead Scoring

**Step 1: Custom Fields**
Create custom fields on Lead and Contact objects:
- Lead_Score__c (Number, 2 decimal places)
- Demographic_Score__c (Number)
- Behavioral_Score__c (Number)
- Score_Last_Updated__c (Date/Time)

**Step 2: Process Builder/Flow**
Build processes to update scores:
- Lead/Contact creation
- Activity logging
- Email engagement
- Website activity (via Pardot/Marketing Cloud)

**Step 3: Lead Assignment Rules**
Update lead assignment rules to consider lead scores:
- High scores to senior reps
- Medium scores to standard queue
- Low scores to nurturing campaigns

### Pipedrive Lead Scoring

**Step 1: Custom Fields**
Add custom fields:
- Lead Score (Numeric)
- Fit Score (Dropdown: A+, A, B+, B, C+, C, D)
- Last Scored (Date)

**Step 2: Automation**
Use Pipedrive automation to:
- Update scores based on activities
- Move high-scoring leads to sales pipeline
- Trigger email sequences for different score ranges

## Advanced Scoring Techniques

### Predictive Lead Scoring

For organizations with substantial historical data, implement machine learning-based scoring:

**Data Requirements**
- 1000+ historical leads
- 100+ conversions
- 12+ months of activity data
- Clean data with outcome labels

**Algorithm Options**
- Logistic Regression (interpretable, works with small data)
- Random Forest (handles missing data well)
- XGBoost (high accuracy, feature importance)
- Neural Networks (for complex patterns)

**Implementation Steps**
1. Data preparation and feature engineering
2. Model training and validation
3. Score calibration (convert to 0-100 scale)
4. Integration with CRM platform
5. Ongoing model monitoring and retraining

### Account-Based Scoring

For B2B companies using account-based marketing:

**Account-Level Scoring**
- Company demographic fit: 40%
- Account engagement breadth: 30%
- Buying committee engagement: 20%
- Intent signals: 10%

**Multi-Contact Scoring**
- Primary contact score (weighted 40%)
- Secondary contacts (weighted 30%)
- Influencer contacts (weighted 20%)
- User-level contacts (weighted 10%)

### Dynamic Scoring Adjustments

**Seasonal Adjustments**
- Increase scoring during peak buying seasons
- Adjust for industry-specific cycles
- Account for economic conditions
- Modify for competitive landscape changes

**Campaign-Specific Scoring**
- Boost scores for specific campaign participants
- Adjust based on campaign performance
- Apply temporary scoring lifts for promotions
- Account for event-driven engagement

## Monitoring and Optimization

### Key Metrics to Track

**Model Performance Metrics**
- Precision: Percentage of high-scored leads that convert
- Recall: Percentage of conversions caught by scoring
- F1 Score: Harmonic mean of precision and recall
- ROC AUC: Overall model discrimination ability

**Business Impact Metrics**
- MQL to SQL conversion rate by score range
- Sales cycle length by lead score
- Deal size correlation with lead score
- Revenue attribution by scored leads

### A/B Testing Framework

**Test Scenarios**
- Different point allocations
- Scoring thresholds for MQL/SQL
- Decay rate variations
- New scoring attributes

**Testing Protocol**
1. Split leads randomly into control/test groups
2. Apply different scoring models
3. Measure conversion rates over 90 days
4. Statistical significance testing (95% confidence)
5. Implement winning variation

### Continuous Improvement Process

**Monthly Reviews**
- Score distribution analysis
- False positive/negative identification
- Sales feedback incorporation
- Performance metric updates

**Quarterly Model Updates**
- Retrain predictive models
- Adjust point allocations
- Update demographic criteria
- Refine behavioral weightings

**Annual Scoring Overhaul**
- Complete customer profile analysis
- Market condition assessment
- Competitive landscape review
- Technology stack evaluation

## Common Pitfalls and Solutions

### Pitfall 1: Over-Complicated Models

**Problem**: Too many variables make the model hard to understand and maintain.

**Solution**: Start with 10-15 key variables that explain 80% of conversions. Add complexity gradually based on performance improvements.

### Pitfall 2: Static Scoring

**Problem**: Scoring models that never change become less accurate over time.

**Solution**: Implement automated decay, regular review cycles, and feedback loops from sales teams.

### Pitfall 3: Ignoring Data Quality

**Problem**: Poor data quality leads to inaccurate scoring and bad decisions.

**Solution**: Implement data validation rules, regular cleaning processes, and progressive profiling strategies.

### Pitfall 4: Not Aligning with Sales

**Problem**: Scoring criteria don't match what sales teams know converts.

**Solution**: Regular collaboration sessions, feedback mechanisms, and joint optimization efforts.

### Pitfall 5: Focusing Only on Demographics

**Problem**: Demographic-only scoring misses engaged prospects who don't fit the "ideal" profile.

**Solution**: Balance demographic fit with behavioral engagement and intent signals.

## Getting Started Checklist

### Week 1: Foundation
- [ ] Define ideal customer profile
- [ ] Analyze historical conversion data
- [ ] Interview sales team on lead quality
- [ ] Set up basic scoring properties in CRM

### Week 2: Model Design
- [ ] Create initial scoring matrix
- [ ] Design demographic scoring criteria
- [ ] Define behavioral scoring rules
- [ ] Set MQL/SQL thresholds

### Week 3: Implementation
- [ ] Build scoring workflows/automation
- [ ] Create lead scoring reports/dashboards
- [ ] Set up decay rules
- [ ] Train team on new process

### Week 4: Testing and Refinement
- [ ] Test scoring on sample leads
- [ ] Validate score accuracy with sales
- [ ] Adjust point allocations
- [ ] Document final model

### Month 2-3: Optimization
- [ ] Monitor conversion rates by score
- [ ] Gather sales feedback
- [ ] Adjust thresholds based on performance
- [ ] Implement advanced features

## Integration with Marketing Automation

### Email Marketing Integration

**Campaign Scoring**
- Segment campaigns by lead score ranges
- Personalize content based on scoring
- Adjust send frequency by engagement level
- Track score changes from email activity

**Drip Campaign Triggers**
- High score leads → immediate sales handoff
- Medium score leads → nurturing sequences
- Low score leads → educational content
- Negative scores → re-engagement campaigns

### Content Marketing Integration

**Dynamic Content Display**
- Show pricing for high-scored visitors
- Display case studies for medium scores
- Offer educational content for low scores
- Customize CTAs based on scoring

**Content Scoring Impact**
- Track which content drives highest scores
- Optimize content for scoring criteria
- Create score-specific content paths
- Measure content ROI by score attribution

## Advanced CRM Integration

### Salesforce Integration

Use Salesforce's Einstein Lead Scoring for enhanced capabilities:
- Automatic model training and updates
- Score explanation features
- Integration with Sales Cloud Einstein
- Advanced reporting and analytics

### HubSpot Integration

Leverage HubSpot's predictive lead scoring:
- Machine learning-based scoring
- Automatic model optimization
- Integration with marketing workflows
- Advanced attribution reporting

### Custom API Integration

For advanced users, build custom scoring systems:
- Real-time scoring updates
- External data source integration
- Custom algorithm implementation
- Advanced analytics and reporting

## Conclusion

Effective lead scoring transforms marketing and sales performance by focusing efforts on the most promising prospects. Start with a simple model based on your ideal customer profile and engagement patterns, then evolve toward more sophisticated approaches as you gather data and experience.

Remember: the best lead scoring system is one that your team actually uses and trusts. Focus on accuracy, simplicity, and continuous improvement rather than complexity.

The tools and templates in this skill will help you implement professional-grade lead scoring that drives real business results. Start with the basics, measure everything, and optimize based on what you learn.