Overview
Score and qualify sales leads to prioritize outreach and focus on the highest-value opportunities. This skill evaluates leads against configurable criteria such as BANT (Budget, Authority, Need, Timeline), company fit, engagement signals, and custom scoring rules. Produces ranked lead lists with scores, recommendations, and next-step suggestions.
Instructions
When a user asks to qualify, score, or prioritize their leads, follow these steps:
Step 1: Load the lead data
Accept leads from CSV, spreadsheet, or manual input:
python
import pandas as pd
def load_leads(file_path):
if file_path.endswith('.csv'):
df = pd.read_csv(file_path)
elif file_path.endswith('.xlsx'):
df = pd.read_excel(file_path)
else:
raise ValueError("Supported formats: CSV, XLSX")
df.columns = [col.strip().lower().replace(' ', '_') for col in df.columns]
return df
Key fields to look for or request:
- Company name, industry, company size (employees or revenue)
- Contact name, title, role
- Lead source (inbound, referral, outbound, event)
- Budget information (if available)
- Current solution / competitor
- Engagement signals (website visits, content downloads, demo requests)
- Notes or qualification details
Step 2: Select the scoring framework
BANT Scoring (default for most B2B):
| Criteria | Weight | Scoring |
|---|---|---|
| Budget | 30% | Has budget (30), exploring (15), no budget (0) |
| Authority | 25% | Decision maker (25), influencer (15), researcher (5) |
| Need | 25% | Urgent need (25), acknowledged need (15), no clear need (5) |
| Timeline | 20% | Within 3 months (20), 3-6 months (10), 6+ months (5) |
Company Fit Scoring (for ICP matching):
| Criteria | Weight | Scoring |
|---|---|---|
| Industry match | 25% | Target industry (25), adjacent (15), other (5) |
| Company size | 25% | Ideal range (25), close (15), outside (5) |
| Geography | 15% | Target region (15), serviceable (10), other (5) |
| Technology fit | 20% | Uses compatible stack (20), neutral (10), incompatible (0) |
| Revenue potential | 15% | High ACV (15), medium (10), low (5) |
Step 3: Score each lead
Match keywords as whole words. Plain substring tests misfire: 'cto' in 'contractor' is true, and '3 month' in '13 months' is true.
python
import re
def has(text, words):
return re.search(r"\b(?:" + "|".join(re.escape(w) for w in words) + r")\b", text, re.I) is not None
def score_lead_bant(lead):
score, reasons = 0, []
# Budget (30 points)
budget = str(lead.get("budget", "")).lower()
if budget in ["yes", "confirmed", "approved"]:
score += 30; reasons.append("Budget confirmed")
elif budget in ["exploring", "pending", "maybe"]:
score += 15; reasons.append("Budget being explored")
else:
reasons.append("No budget information")
# Authority (25 points)
title = str(lead.get("title", ""))
if has(title, ["ceo", "cto", "cfo", "coo", "founder", "owner", "vp", "svp", "evp", "vice president", "director", "head of"]):
score += 25; reasons.append("Decision maker")
elif has(title, ["manager", "lead", "senior", "principal"]):
score += 15; reasons.append("Influencer role")
else:
score += 5; reasons.append("Researcher / early stage")
# Need (25 points)
need = str(lead.get("need", lead.get("pain_point", ""))).strip().lower()
if has(need, ["urgent", "critical", "immediate", "asap"]):
score += 25; reasons.append("Urgent need identified")
elif need and need not in ["none", "no", "nan"]:
score += 15; reasons.append("Need acknowledged")
else:
score += 5; reasons.append("Need unclear")
# Timeline (20 points)
timeline = str(lead.get("timeline", "")).lower()
months = re.search(r"\b(\d+)\s*(?:-\s*\d+\s*)?months?\b", timeline) # "3 months", "4-6 months"
if has(timeline, ["now", "asap", "this month", "this quarter"]) or (months and int(months.group(1)) <= 3):
score += 20; reasons.append("Near-term timeline")
elif has(timeline, ["next quarter", "this year"]) or (months and int(months.group(1)) <= 6):
score += 10; reasons.append("Medium-term timeline")
else:
score += 5; reasons.append("No clear timeline")
return score, reasons
MEDDIC (for larger, multi-stakeholder deals) scores six questions instead: Metrics (the quantified business result), Economic buyer (who signs), Decision criteria, Decision process, Identify pain, Champion (an internal advocate). Many teams add a seventh C for Competition (MEDDICC). Give each a 0-2 rating (unknown, partial, confirmed) and rank by the total; unanswered questions are the next discovery-call agenda.
Step 4: Classify and rank leads
Assign qualification tiers based on score:
python
def classify_lead(score):
if score >= 80:
return "Hot", "Immediate follow-up - schedule a call today"
elif score >= 60:
return "Warm", "Priority follow-up within 48 hours"
elif score >= 40:
return "Nurture", "Add to nurture campaign, follow up in 1-2 weeks"
else:
return "Cold", "Low priority - monitor for engagement changes"
Step 5: Generate the qualification report
Output a ranked list with scores, tiers, reasons, and recommended actions:
python
def qualify(df, output_path="qualified_leads.csv"):
rows = []
for lead in df.to_dict("records"):
score, reasons = score_lead_bant(lead)
tier, action = classify_lead(score)
rows.append({**lead, "score": score, "tier": tier, "action": action, "reasons": "; ".join(reasons)})
ranked = pd.DataFrame(rows).sort_values("score", ascending=False)
ranked.to_csv(output_path, index=False)
return ranked
Examples
Example 1: Qualify a list of inbound leads
User request: "Score these 25 leads from our webinar sign-ups. Here's the CSV file."
Actions taken:
- Load leads.csv with 25 entries
- Apply BANT scoring based on available fields
- Rank and classify all leads
Output:
Lead Qualification Report
=========================
Total leads scored: 25
Tier Breakdown:
Hot (80+): 4 leads (16%) - Immediate follow-up
Warm (60-79): 7 leads (28%) - Priority follow-up
Nurture (40-59): 9 leads (36%) - Nurture campaign
Cold (<40): 5 leads (20%) - Monitor
Top 5 Leads:
Rank | Company | Contact | Score | Tier | Action
1 | TechGrowth Inc | Sarah Chen, VP| 95 | Hot | Call today - budget confirmed, urgent need
2 | DataFirst LLC | Mark Jones, Dir| 90 | Hot | Call today - decision maker, Q1 timeline
3 | ScaleUp Corp | Amy Park, CTO | 85 | Hot | Call today - strong tech fit, active eval
4 | CloudNine SaaS | Tom Lee, Head | 80 | Hot | Demo this week - exploring budget
5 | GreenField Co | Lisa Wang, Mgr| 75 | Warm | Follow up in 48h - need confirmed
Full report saved: qualified_leads.csv
Example 2: Evaluate leads against ideal customer profile
User request: "We sell to mid-market SaaS companies (100-1000 employees) in North America. Score these leads for fit."
Actions taken:
- Define ICP criteria from user description
- Score each lead on company fit dimensions
- Highlight best-fit and poor-fit leads
Output:
ICP Fit Analysis
================
Target Profile: Mid-market SaaS, 100-1000 employees, North America
Fit Distribution:
Strong fit (80+): 8 leads - Match ICP on 4+ criteria
Moderate fit (60-79): 12 leads - Match on 2-3 criteria
Weak fit (<60): 5 leads - Poor ICP match
Strong Fit Leads:
1. StreamLine (SaaS, 450 emp, US) - Score: 92
2. DataPipe (SaaS, 280 emp, Canada) - Score: 88
3. FlowMetrics (SaaS, 620 emp, US) - Score: 85
Weak Fit (consider deprioritizing):
- MegaCorp Industries (Manufacturing, 15,000 emp) - Score: 25
- LocalShop (Retail, 12 emp, UK) - Score: 20
Example 3: Pipeline quality analysis
User request: "Analyze our current pipeline of 50 opportunities and tell me which ones to focus on this quarter"
Actions taken:
- Load pipeline data
- Score on BANT plus deal velocity indicators
- Identify stalled deals and top opportunities
Output:
Pipeline Quality Analysis: Q4 2026
====================================
Total opportunities: 50
Total pipeline value: $2.4M
Quality Distribution:
High quality (likely to close): 12 deals ($820K)
Medium quality (needs work): 18 deals ($740K)
Low quality (at risk): 14 deals ($580K)
Stalled (no activity 30+ days): 6 deals ($260K)
Recommended Focus (Top 10):
1. Northwind Logistics - $120K - score 92 - sign-off meeting with CFO booked for Nov 4
2. Brightline Health - $85K - score 88 - security review pending, send SOC 2 report
(8 more, each with score and next step)
Stalled Deals Requiring Attention:
- Harbor Freight Systems ($60K): no contact in 45 days, re-engage or close out
- Lumen Analytics ($45K): champion left the company, identify a new sponsor
Quarterly Forecast:
Best case: $820K (high quality deals)
Likely case: $620K (weighted by probability)
Worst case: $380K (only committed deals)
Guidelines
- Default to BANT scoring for B2B leads unless the user specifies a different framework.
- Always sort output by score descending so the best leads appear first.
- Include a specific recommended action for each lead tier, not just the score.
- When lead data is incomplete, score based on available fields and note what information is missing.
- Flag leads where the contact title suggests they are not a decision maker but the deal is otherwise strong.
- For pipeline analysis, flag stalled deals (no activity in 30+ days) as a separate category.
- Scoring weights should be configurable. Use the defaults but adjust if the user specifies different priorities.
- Never present raw scores without context. Always include the tier classification and a recommended action.
- Install in a virtual environment:
python -m venv .venv && .venv/bin/pip install pandas openpyxl(pandas 3 needs Python 3.11+). - Lead data is personal data (names, work emails): keep exports out of shared folders and version control, and do not send it to third-party enrichment tools without the user's approval.
- Scores rank leads; they do not predict closing. Revisit the weights after comparing scores with real won/lost outcomes.