--- name: lead-scoring-model description: Design and implement lead scoring models with demographic, behavioral, and negative scoring rules, MQL/SQL threshold calibration, score decay mechanics, lead routing automation, and predictive scoring integration. Use when building lead scoring systems, defining MQL criteria, creating behavioral scoring rules, optimizing lead routing, calibrating score thresholds, or aligning marketing and sales on lead quality. Triggers on phrases like "lead scoring", "MQL definition", "SQL definition", "lead quality", "scoring model", "marketing qualified lead", "sales qualified lead", "lead routing", "lead prioritization", "score threshold", "predictive scoring", "intent scoring", "lead recycling", "score decay". --- # Lead Scoring Model Designer Build lead scoring frameworks that help sales teams focus on the highest-quality, sales-ready leads, with systematic calibration, routing automation, and continuous optimization. ## Scoring Model Architecture ``` LEAD SCORING MODEL DESIGN ═══════════════════════════════════════════════════════════ Model Type Selection: → Fit-Based Scoring (Static): Who they are (firmographics, demographics) - Best for: Clear ICP, predictable buyer profiles - Limitation: Doesn't capture buying intent → Behavior-Based Scoring (Dynamic): What they do (engagement, intent signals) - Best for: Content-rich marketing, multi-touch journeys - Limitation: High-volume noise without fit filter → Combined Scoring (Recommended): Fit × Behavior = Total Score - Formula: Total Score = (Fit Score × 0.40) + (Behavior Score × 0.60) - Best for: Most B2B organizations - Effect: Lead must be BOTH right fit AND showing intent → Predictive Scoring (AI/ML): Historical deal outcomes train model - Best for: Organizations with 500+ closed deals (sufficient training data) - Tools: 6sense, MadKitti, Salesforce Einstein, HubSpot Predictive Lead Scoring - Accuracy: 20–40% improvement over manual scoring (when trained on quality data) SCORE DECAY MECHANICS (Critical for stale lead management): → Purpose: Reduce score of inactive leads over time (prevents "zombie" MQLs) → Decay schedule: - No activity 30 days: -10 points - No activity 60 days: -20 points - No activity 90 days: -30 points - No activity 120 days: -50 points (effectively resets lead) → Decay exceptions (do NOT decay): - Lead engaged with sales team (active opportunity in CRM) - Lead in active nurture sequence (email engagement) - Lead from high-value account (ABM target) → Net effect: Lead score reflects recency AND relevance FIRMOGRAPHIC / DEMOGRAPHIC SCORES (Fit — Weight: 40%): ┌─────────────────────────────────┬────────────┬────────────────────────────────────┐ │ Criteria │ Points │ Notes │ ├─────────────────────────────────┼────────────┼────────────────────────────────────┤ │ COMPANY SIZE (employees) │ │ Align with your ICP │ │ 1–50 │ 5 │ SMB — may be self-serve │ │ 51–200 │ 10 │ Small business │ │ 201–1,000 │ 15 │ Mid-market (if ICP) │ │ 1,001–5,000 │ 15 │ Mid-market/lower enterprise │ │ 5,001–10,000 │ 10 │ Enterprise (if ICP) │ │ 10,000+ │ 5 │ May be too large (complex sales) │ │ │ │ │ │ ANNUAL REVENUE │ │ Adjust based on your ACV │ │ <$1M │ 5 │ │ │ $1M–$10M │ 10 │ │ │ $10M–$50M │ 15 │ │ │ $50M–$200M │ 15 │ │ │ $200M+ │ 10 │ │ │ │ │ │ │ INDUSTRY │ │ │ │ Primary target industry │ 15 │ e.g., SaaS, Healthcare, FinTech │ │ Secondary target industry │ 10 │ e.g., Manufacturing, Retail │ │ Non-target industry │ 0 │ │ │ Restricted industry │ -50 │ e.g., Government, Cannabis │ │ │ │ │ │ JOB TITLE / ROLE │ │ │ │ Economic buyer (C-suite, VP) │ 20 │ Budget authority │ │ Influencer (Director, Manager) │ 15 │ Decision influence │ │ End user (IC, Analyst) │ 10 │ Can champion but not buy │ │ Student / intern │ -50 │ Disqualify │ │ Competitor employee │ -100 │ Auto-disqualify │ │ │ │ │ │ GEOGRAPHY │ │ │ │ Primary target region │ 10 │ e.g., North America, Western EU │ │ Secondary target region │ 5 │ e.g., Asia-Pacific, LATAM │ │ Non-target region │ 0 │ │ │ Restricted region │ -50 │ Compliance/legal restrictions │ │ │ │ │ │ TECHNOLOGY STACK │ │ Use BuiltWith/LinkedIn for data │ │ Uses complementary tech │ 10 │ Integration opportunity │ │ Uses competitor tech │ 5 │ Displacement opportunity │ │ Uses outdated tech │ 15 │ Migration opportunity │ │ │ │ │ │ FIT SCORE RANGE │ 0–100 │ Target MQL fit score: ≥ 40 │ └─────────────────────────────────┴────────────┴────────────────────────────────────┘ BEHAVIORAL SCORES (Intent — Weight: 60%): ┌─────────────────────────────────┬────────────┬────────────────────────────────────┐ │ Activity │ Points │ Notes │ ├─────────────────────────────────┼────────────┼────────────────────────────────────┤ │ EMAIL ENGAGEMENT │ │ Per-email event │ │ Email opened │ +2 │ Cap: 6 points (3 opens) │ │ Email link clicked │ +5 │ Cap: 15 points (3 clicks) │ │ Email replied │ +10 │ Strong intent signal │ │ Email marked spam │ -100 │ Auto-disqualify │ │ Email unsubscribed │ -100 │ Auto-disqualify │ │ │ │ │ │ WEBSITE ACTIVITY │ │ Per-page-view event │ │ Homepage visit │ +2 │ Low intent │ │ Blog post read │ +2 │ Cap: 10 points (5 posts) │ │ Product feature page │ +5 │ Moderate intent │ │ Pricing page visit │ +10 │ High intent │ │ Case study page │ +5 │ Social proof interest │ │ Demo/trial page │ +10 │ High intent │ │ Multiple pages per session │ +5 │ Deep engagement │ │ Return visitor (3+ visits) │ +10 │ Sustained interest │ │ │ │ │ │ CONTENT ENGAGEMENT │ │ Per-download event │ │ Blog comment │ +3 │ Active engagement │ │ Checklist/template download │ +5 │ Practical intent │ │ White paper download │ +10 │ Research phase │ │ Case study download │ +8 │ Evaluation phase │ │ ROI calculator use │ +10 │ High intent — evaluating value │ │ Product comparison download │ +12 │ Active evaluation │ │ │ │ │ │ EVENT ENGAGEMENT │ │ │ │ Webinar registered │ +5 │ │ │ Webinar attended (live) │ +10 │ │ │ Webinar replay watched │ +5 │ │ │ Trade show booth visit │ +10 │ │ │ Event session attended │ +8 │ │ │ │ │ │ │ SALES / PRODUCT INTERACTION │ │ Strongest intent signals │ │ Website chat initiated │ +10 │ │ │ Support ticket created │ +5 │ │ │ Free trial started │ +15 │ Very high intent │ │ Demo requested │ +20 │ AUTO-MQL trigger │ │ Added items to cart │ +10 │ │ │ Pricing page + 3+ clicks │ +15 │ AUTO-MQL trigger │ │ │ │ │ │ BEHAVIOR SCORE RANGE │ 0–100 │ Target MQL behavior score: ≥ 40 │ └─────────────────────────────────┴────────────┴────────────────────────────────────┘ NEGATIVE / DISQUALIFICATION SCORES: → Email bounced (hard): -100 (remove from active lists immediately) → Marked as spam: -100 (remove from all lists, review send practices) → Unsubscribed: -100 (compliance requirement — honor immediately) → Student email (.edu): -50 (not a buyer) → Competitor domain: -100 (protect competitive intelligence) → Government/restricted: -50 (compliance/legal restrictions) → Job title = intern/student/volunteer: -50 (not a buyer) → Company = non-profit (if not target): -50 → Any negative score applied: Lead enters disqualification review ``` ## MQL/SQL Thresholds & Calibration ``` MQL/SQL DEFINITIONS AND THRESHOLDS ═══════════════════════════════════════════════════════════ MQL (Marketing Qualified Lead) — Ready for Sales Outreach: → Score threshold: Total score ≥ 60 (out of 100) (Fit score ≥ 40 AND Behavior score ≥ 40) → Auto-MQL triggers (bypass score threshold): - Demo request submitted - Free trial started - Pricing page visit + 3+ additional page clicks in same session - ROI calculator used - Contacted sales directly (phone, email, chat) → MQL handoff process: - Marketing generates MQL → Salesforce/HubSpot workflow triggers - Lead assigned to rep within 5 minutes (SLA) - Rep has 24 hours to first contact (email or call) - If no contact in 24 hours → reassign to next available rep SQL (Sales Qualified Lead) — Ready for Opportunity Creation: → SQL criteria (ALL must be met): 1. Lead meets MQL threshold (score ≥ 60) 2. Rep has completed discovery call (or BANT phone qualification) 3. Confirmed: Budget exists or can be allocated 4. Confirmed: Timeline defined (purchase within 90 days) 5. Confirmed: Decision process understood 6. Confirmed: No disqualification factors → SQL conversion rate target: 20–30% of MQLs become SQLs (Below 15%: MQL threshold too low, sending unqualified leads) (Above 35%: MQL threshold too high, missing qualified leads) OPPORTUNITY (Oppty-Qualified Lead) — Active Deal in Pipeline: → Opportunity criteria (ALL must be met): 1. Lead meets SQL criteria 2. Demo completed with prospect 3. Economic buyer identified (and engaged or planned) 4. Quantified business impact captured 5. Competitive situation understood 6. Forecast close date set → Opportunity conversion rate target: 40–60% of SQLs become Opportunities (Below 30%: SQL criteria too loose, sending poor leads to demo) (Above 70%: SQL criteria too tight, missing viable deals) CALIBRATION METHODOLOGY (Monthly): → Pull historical data: Last 90 days of leads, by score range → Calculate conversion rates: MQL → SQL → Opportunity → Closed-Won, by score range → Adjust thresholds based on data: ┌────────────────┬──────────┬──────────┬──────────┬──────────────────┐ │ Score Range │ Leads │ SQL Rate │ Opp Rate │ Win Rate (of Opp)│ ├────────────────┼──────────┼──────────┼──────────┼──────────────────┤ │ 80–100 │ 150 │ 35% │ 55% │ 30% │ │ 60–79 │ 300 │ 25% │ 45% │ 25% │ │ 40–59 │ 400 │ 15% │ 30% │ 18% │ │ 20–39 │ 350 │ 8% │ 15% │ 10% │ │ 0–19 │ 200 │ 3% │ 5% │ 5% │ └────────────────┴──────────┴──────────┴──────────┴──────────────────┘ → Insight: Leads scoring 60+ have 2.5x higher SQL rate than 40–59 → Action: MQL threshold of 60 is validated (25% SQL rate = on target) → Action: Leads scoring 40–59 → add to nurture track, not MQL queue CALIBRATION ADJUSTMENT RULES: → If SQL rate for MQLs < 15%: Raise MQL threshold by 5 points → If SQL rate for MQLs > 35%: Lower MQL threshold by 5 points → If win rate for 80+ leads < 20%: Review high-score criteria (false positives) → If win rate for 60–79 leads > 25%: Consider raising minimum threshold → Recalibrate quarterly minimum, monthly if significant changes to ICP or offering ``` ## Lead Routing & SLA Management ``` LEAD ROUTING RULES ═══════════════════════════════════════════════════════════ Routing Strategy (select ONE primary, ONE secondary): 1. Territory-Based Routing (Primary): → By region: North America → NA team, EMEA → EMEA team, APAC → APAC team → By state/province: CA → West Coast team, NY → East Coast team → By postal code: Automated lookup, assigned to rep with matching territory → Tools: Salesforce Territory Management, HubSpot Lead Routing 2. Industry-Based Routing (Primary): → Healthcare → Healthcare vertical rep → Financial Services → FinTech vertical rep → Technology → Tech vertical rep → Mixed industries → Round-robin within generalist team → Tools: Custom Salesforce/HubSpot routing rules 3. Round-Robin Routing (Secondary/Fallback): → Distribute evenly among available reps → Respect capacity limits (max 50 active MQLs per rep) → Skip unavailable reps (on PTO, out of office) → Tools: Salesforce Round-Robin, HubSpot Lead Rotation 4. Capacity-Based Routing (Advanced): → Assign to rep with lowest current workload → Consider rep's current pipeline value and deal count → Tools: Clari, Salesforce Einstein Activity Capture 5. Score-Priority Routing (Premium): → Highest-score leads → Most experienced/senior reps → Lower-score leads → Junior reps or SDR team → Tools: Custom routing rules in CRM RESPONSE TIME SLA: ┌────────────────────────────────┬────────────┬──────────────────────────────┐ │ SLA Tier │ Time │ Expected Impact │ ├────────────────────────────────┼────────────┼──────────────────────────────┤ │ Gold (Score ≥ 80 or Auto-MQL) │ 5 minutes │ 6x more likely to convert │ │ Silver (Score 60–79) │ 1 hour │ 3x more likely to convert │ │ Bronze (Score 40–59, nurture) │ 24 hours │ Baseline conversion │ └────────────────────────────────┴────────────┴──────────────────────────────┘ → Research: Leads contacted within 5 minutes are 21x more likely to convert (source: InsideSales/Engage study) → Monitoring: Weekly SLA compliance report (reps, managers, marketing) → Escalation: If rep misses SLA, auto-reassign to next available rep LEAD RECYCLING PROCESS: → Near-MQL leads (Score 40–59): - Enter 60-day nurture sequence (3 emails/week, educational content) - Re-score after 60 days (if score reaches 60+, enter MQL queue) - If score < 40 after 60 days: Enter long-term nurture (1 email/week) → Engaged but not-ready leads (Score 60+, timeline > 90 days): - Enter quarterly nurture sequence - Re-engage on trigger events (funding, hiring, product launches) - Score decays over time (see decay mechanics) → Disqualified leads (negative score or explicit opt-out): - Archive in CRM (Status: Disqualified) - Do NOT re-enter active lists - Exception: Re-evaluate after 12 months if ICP changed LEAD RECYCLING FUNNEL: ┌───────────────────────────────────┬──────────────────────────────────┐ │ Lead Status │ Action │ ├───────────────────────────────────┼──────────────────────────────────┤ │ MQL → Rep responded, not SQL │ Enter 30-day nurture │ │ (score ≥ 60, not qualified) │ Re-score at 30 days │ ├───────────────────────────────────┼──────────────────────────────────┤ │ SQL → Deal progressed, lost │ Enter 90-day nurture │ │ (qualified, lost to competitor) │ Re-engage at competitor renewal │ ├───────────────────────────────────┼──────────────────────────────────┤ │ SQL → Deal progressed, timing │ Enter 60-day nurture │ │ (qualified, timing issue) │ Re-engage at next budget cycle │ ├───────────────────────────────────┼──────────────────────────────────┤ │ Near-MQL → score 40-59 │ Enter 60-day nurture │ │ (close to qualified) │ Re-score at 60 days │ └───────────────────────────────────┴──────────────────────────────────┘ ``` ## Integration Points - **Salesforce / HubSpot**: Lead scoring configuration, MQL/SQL workflows, automated lead routing, score tracking, SLA monitoring — $45–$3,600/month - **6sense / MadKitti**: AI-powered intent scoring, predictive lead scoring, account-based scoring, anomaly detection — $10,000–$50,000/year - **Demandbase / Terminus**: ABM platform, account scoring, multi-touch intent tracking, program management — $25,000–$100,000/year - **Segment / mParticle**: Customer data platform, behavioral event tracking, real-time scoring data pipeline — $5,000–$20,000/year - **Apollo / ZoomInfo**: Data enrichment (firmographics, technographics), lead scoring data accuracy — $49–$800/month - **Salesforce Einstein / HubSpot Predictive**: ML-based lead scoring, auto-generated score recommendations, conversion probability — included in enterprise plans - **Tableau / Looker**: Lead score analytics dashboards, conversion rate tracking, score calibration reports — $30–$75/user/month ## Edge Cases - **Account-Based Marketing (ABM)** — Score accounts, not just contacts: → Account-level scoring: Aggregate individual contact scores within account → Threshold trigger: Account qualifies when 3+ contacts reach individual MQL → Account score formula: Sum of top 5 contact scores × 0.6 + Account firmographic score × 0.4 → Routing: Assigned to enterprise AE (not SDR) based on account score and strategic value → Engagement tracking: Multiple touchpoints across account employees trigger score increases → Example: Acme Corp has 5 employees who opened emails, 2 visited pricing page, 1 requested demo → Account score: (15+12+10+8+5) × 0.6 + 80 × 0.4 = 65 → ABM program activation - **PLG / Freemium Models** — Score based on product usage: → Usage-based scoring (product events = behavioral scores): - Signup: +10 points - First key action completed: +15 points - Weekly active use (3+ days/week): +10 points - Team invite sent: +10 points (expansion signal) - Reached usage threshold (e.g., 50% of plan limit): +15 points (upgrade signal) - Inactive 14+ days: -10 points (churn risk) → SQL trigger: Usage-based activation metric (e.g., "completed onboarding + 3 team members") → Conversion: 5–15% of free users convert to paid (industry benchmark) → Tools: Mixpanel, Amplitude, Pendo for usage tracking - **Low deal volume** (< 50 closed deals/year — insufficient calibration data): → Use industry benchmarks for initial score thresholds (not historical data) → Focus on qualitative feedback from sales reps ("what makes a good lead?") → Revisit scoring model every 6 months until sufficient data accumulates → Minimum data for reliable calibration: 200+ leads tracked through full funnel → Supplement with competitor and industry data (G2, Capterra, PeerSpark) - **Score inflation** (all leads scoring high, losing discrimination): → Symptom: MQL SQL rate drops below 15% (too many "MQLs" are actually unqualified) → Cause: Point values too generous, too many activities tracked → Fix: Recalibrate point values to create wider score distribution → Target distribution: - Score 80–100: 10–15% of leads (premium, immediate action) - Score 60–79: 20–30% of leads (MQL, standard routing) - Score 40–59: 30–40% of leads (near-MQL, nurture) - Score 20–39: 15–25% of leads (long-term nurture) - Score 0–19: 5–10% of leads (disqualified or very early) → Test: Run scoring model on 1,000 leads and check distribution - **Multi-touch attribution** (lead engaged across multiple channels): → Challenge: Lead opened email, visited website, downloaded content, attended webinar Which touchpoint gets credit for score increase? → Solution: Incremental scoring (each touch adds points, no cap per category) → Advanced: Time-weighted scoring (recent activities worth more than older ones) - Activity in last 7 days: Full points - Activity 8–30 days ago: 75% of points - Activity 31–60 days ago: 50% of points - Activity 61+ days ago: 25% of points → Implementation: CRM or CDP calculates time-weighted score automatically - **International/global scoring** (multiple regions with different buying behaviors): → Region-specific scoring weights (different behaviors matter differently by market): - North America: Email engagement heavily weighted (40% of behavior score) - Europe: Content downloads heavily weighted (50% of behavior score) — GDPR-conscious - Asia-Pacific: Website visits heavily weighted (50% of behavior score) — lower email engagement → Currency considerations: Pricing page visits weighted differently (price sensitivity varies) → Language: Score multilingual content engagement separately → Compliance: Ensure scoring doesn't use protected characteristics (GDPR, CCPA)