Auto-Score WhatsApp Leads with AI Agents

Your sales team wastes 60% of their time on leads that will never buy. AI lead scoring sends hot leads to sales instantly and nurtures cold ones automatically. Stop wasting your best people on dead-end conversations.

MoltFlow's lead scoring runs on YOUR criteria, not a black-box algorithm. You define what "hot" means for your business. Sales teams using AI lead scoring spend 60% more time with qualified buyers.

This guide shows you how to build an automated lead qualification system using AI agents and MoltFlow's lead detection. Your system will score every WhatsApp lead in real-time, identify high-intent prospects instantly, and route them to your sales team while they're most engaged.

What You'll Need

Before starting, make sure you have:

• A MoltFlow account with Business plan or higher — Lead scoring requires the Business plan (stop paying sales reps to qualify tire-kickers). Compare plans here.
• OpenClaw API key — Get your API key from OpenClaw's developer portal. You'll need the Pro tier for production lead volumes (10,000 messages/month).
• At least one monitored WhatsApp group — This could be a community group, product interest group, or any group where potential customers gather. Make sure you have admin permissions.
• Clear ideal customer profile (ICP) — Before building your scoring system, document what makes a qualified lead for your business: budget signals, use case fit, urgency indicators, decision-making authority.

Important: Ensure you comply with WhatsApp's terms of service regarding group message monitoring and have appropriate consent from group members if required in your jurisdiction.

Step 1: Understand the Lead Qualification Pipeline

Before building your system, let's understand how the automated pipeline works:

The flow:

1. Group message received → Someone posts in your monitored WhatsApp group
2. Lead detection → MoltFlow identifies messages that indicate purchase interest based on keywords and patterns
3. AI scoring → OpenClaw agent analyzes the message and assigns a qualification score (1-10)
4. Qualification tier assignment → Leads are categorized as Hot (7-10), Warm (4-6), or Cold (1-3)
5. Automated routing → Hot leads trigger immediate notifications to your sales team; warm leads enter nurture sequences; cold leads are logged for later analysis
6. Follow-up execution → Automated messages sent based on lead tier and scoring results

Why this works:

• Speed — Leads are scored within seconds, not hours or days
• Consistency — Every lead is evaluated against the same criteria
• Scale — Your team can monitor dozens of groups without adding headcount
• Focus — Sales reps only engage with qualified, high-intent prospects

MoltFlow's built-in lead detection:

MoltFlow automatically identifies potential leads from monitored groups by detecting:

• Questions about pricing, features, availability
• Expressions of pain points your product solves
• Mentions of competitor products
• Buying signals: "need", "looking for", "urgently", "budget"
• Contact information shared (email, website, LinkedIn)

These detected leads appear in your Leads page in the dashboard, ready for AI scoring.

Step 2: Set Up Group Monitoring for Lead Detection

First, configure which WhatsApp groups MoltFlow should monitor for potential leads. Learn more about monitoring strategies in our group monitoring guide.

Add groups to monitor:

1. Navigate to the Groups page in your MoltFlow dashboard
2. Click "Monitor Group" button
3. Select groups from your WhatsApp account's group list
4. For each group, configure:
   • Group name (for your reference)
   • Lead detection mode: Keyword-based or All messages
   • Priority level: High/Medium/Low (affects scoring later)

Configure keyword triggers:

If you chose keyword-based detection, add triggers that indicate purchase intent:

For B2B SaaS:

• "pricing", "cost", "budget", "demo", "trial", "integrate"
• "looking for", "need help with", "recommendations"
• "compared to [competitor]", "better than [alternative]"

For e-commerce:

• "buy", "order", "purchase", "shipping", "available"
• "discount", "coupon", "deal", "price"
• "out of stock", "when available", "restock"

For professional services:

• "hire", "looking for", "project", "consultation", "quote"
• "budget", "timeline", "urgently need", "ASAP"
• "recommend", "experienced in", "portfolio"

Important: Start with 10-15 high-signal keywords. Too many keywords generate noise; too few miss opportunities. You can refine based on results. See our keyword rules guide for advanced filtering strategies.

Verify lead detection is working:

1. After saving your group monitoring config, wait for new messages to arrive (or have a colleague post a test message with a keyword)
2. Navigate to the Leads page in your dashboard
3. Verify the message appears as a detected lead
4. Check that the lead shows: contact name, phone number, message preview, group source, timestamp

If leads aren't appearing, check:

• Your WhatsApp session is in WORKING state
• You're an admin in the monitored group
• Keyword matching is case-insensitive (test both "Pricing" and "pricing")
• Webhook events are flowing (check Sessions → Events)

Step 3: Configure OpenClaw for Lead Scoring

Now we'll set up OpenClaw to analyze each detected lead and assign a qualification score.

Create a scoring agent in OpenClaw:

1. Navigate to AI Configuration in MoltFlow
2. Add a new OpenClaw configuration specifically for lead scoring
3. Use openclaw-chat-pro model (better reasoning for nuanced scoring)
4. Set the following system prompt:

You are a lead qualification AI for [Your Company Name]. Your job is to analyze WhatsApp messages from potential customers and assign a lead score from 1-10.

SCORING CRITERIA:

1. Purchase Intent (0-3 points):
   - 3 = Explicit intent: "want to buy", "need pricing", "ready to order"
   - 2 = Strong interest: "considering", "comparing options", "tell me more"
   - 1 = Mild interest: "curious", "looking into", "maybe"
   - 0 = No clear intent: general question, off-topic

2. Budget Signals (0-3 points):
   - 3 = Budget confirmed: mentions price range, "have budget", "approved"
   - 2 = Budget implied: asks about pricing, payment plans, ROI
   - 1 = Budget uncertain: "affordable?", "expensive?", "cost?"
   - 0 = No budget signals

3. Urgency (0-2 points):
   - 2 = Immediate: "ASAP", "urgent", "this week", "deadline"
   - 1 = Soon: "in the next month", "planning to", "soon"
   - 0 = No urgency: "eventually", "someday", "thinking about"

4. Fit with Our Product (0-2 points):
   - 2 = Perfect fit: describes exact use case we solve
   - 1 = Potential fit: overlapping needs, some alignment
   - 0 = Poor fit: needs we don't address, wrong segment

RESPONSE FORMAT:

Return ONLY valid JSON (no markdown formatting):

{
  "score": <total score 1-10>,
  "breakdown": {
    "purchase_intent": <0-3>,
    "budget_signals": <0-3>,
    "urgency": <0-2>,
    "product_fit": <0-2>
  },
  "reasoning": "<2-3 sentence explanation of the score>",
  "recommended_action": "<immediate_contact | schedule_follow_up | nurture | disqualify>",
  "suggested_response": "<personalized message to send to this lead>"
}

IMPORTANT:
- Be conservative with scores. Only assign 8+ for clearly qualified, ready-to-buy leads.
- Factor in the conversation context if available (not just a single message).
- If the message is unclear or lacks information, default to a mid-range score (4-5) for human review.

Customize the scoring criteria for your business:

The example above works for most B2B SaaS and service businesses, but you should adapt it:

For e-commerce:

• Replace "purchase intent" with "add to cart likelihood"
• Add "product familiarity" (0-2): knows exactly what they want vs. browsing

For high-ticket B2B:

• Add "decision authority" (0-2): decision maker, influencer, or researcher
• Increase weight on "budget signals" (0-4 instead of 0-3)

For professional services:

• Add "project scope" (0-2): well-defined vs. vague
• Add "timeline clarity" (0-2): specific start date vs. someday

The key is aligning your scoring with what actually predicts closed deals in your business. Analyze your past wins and identify the signals that were present early in the conversation.

Step 4: Define Qualification Criteria and Scoring Rules

Now configure how MoltFlow should act on the AI-generated scores.

Set score thresholds:

1. Navigate to Lead Qualification settings in MoltFlow
2. Define your tier thresholds:

Recommended thresholds:

• Hot leads (7-10 points) — Immediate sales engagement required

  • Assign label: "🔥 Hot Lead"
  • SLA: Contact within 15 minutes
  • Routing: Direct to senior sales rep

• Warm leads (4-6 points) — Qualified but not urgent

  • Assign label: "⚡ Warm Lead"
  • SLA: Contact within 4 hours
  • Routing: Enter automated nurture sequence

• Cold leads (1-3 points) — Low priority or poor fit

  • Assign label: "❄️ Cold Lead"
  • SLA: Review weekly
  • Routing: Add to general newsletter list

Configure label auto-assignment:

MoltFlow can automatically tag leads based on their score:

1. In Lead Qualification settings, enable Auto-Label
2. Map score ranges to labels (as shown above)
3. These labels will appear in your Leads page and on the contact's WhatsApp chat

Set up disqualification rules:

Some messages should skip scoring entirely:

• Spam patterns: repeated messages, link spam, promotional content
• Existing customers: check against your customer database
• Unrelated topics: off-topic messages that triggered keywords by accident

Configure these in Lead Filters → Auto-Disqualify rules.

Step 5: Set Up Automated Follow-Up Sequences

Now configure what happens automatically after a lead is scored. For bulk outreach to segmented lists, see our bulk messaging guide.

For hot leads (7-10) — Immediate personal outreach:

1. Navigate to Scheduled Messages in MoltFlow
2. Create a new message template: "Hot Lead - Immediate Response"
3. Set trigger: When lead score >= 7
4. Message content (personalize with variables):

Hi {contact_name}! 👋

I saw your message in {group_name} about {topic}. I'd love to help you with that!

{suggested_response_from_ai}

I'm available for a quick call in the next hour if you'd like to discuss details. Or we can schedule a time that works better for you.

When would be convenient?

Best,
{sales_rep_name}
{company_name}

Important: Hot leads should get personalized, human-sounding messages immediately. Use the AI's suggested_response to show you understand their specific need.

For warm leads (4-6) — Automated nurture sequence:

Create a 3-message sequence over 7 days:

Day 0 (immediate):

Hi {contact_name}!

Thanks for your interest in {product_name}. I noticed you were asking about {topic} in {group_name}.

I'd be happy to share some resources that might help. Would you like me to send over:
1. A quick overview of how we solve [pain point]
2. Pricing information
3. A case study from a similar customer

Let me know what would be most useful!

Day 3:

Hi {contact_name},

Just following up — did you get a chance to review the info I sent?

I'm here if you have any questions about {product_name}. A lot of our customers had similar questions about {common_objection} when they started — happy to explain how we handle that.

No pressure, just want to make sure you have what you need!

Day 7:

Hi {contact_name},

Last check-in from me! If you're still interested in {product_name}, I'd love to schedule a quick 15-minute demo this week.

If the timing isn't right, no worries — feel free to reach out whenever you're ready.

All the best,
{sales_rep_name}

Use MoltFlow's scheduled message feature to automate this entire sequence.

For cold leads (1-3) — Log and monitor:

Don't send automated follow-ups to cold leads. Instead:

1. Log them in your CRM or customer database
2. Add to a general marketing list (if they've consented)
3. Set up a monthly review process to manually check for any patterns you missed

Organize cold leads into custom groups for future campaigns using our custom groups guide.

Step 6: Route Qualified Leads to Your Sales Team

Set up real-time notifications so your team can act on hot leads immediately.

Option A: WhatsApp notifications to a sales team group

1. Create a dedicated WhatsApp group for your sales team
2. In MoltFlow, go to Webhooks → Create Webhook
3. Set trigger: Lead score >= 7
4. Action: Send message to sales team group
5. Message template:

🔥 HOT LEAD ALERT 🔥

Contact: {contact_name}
Phone: {phone_number}
Score: {lead_score}/10
Source: {group_name}

Message:
"{message_content}"

AI Reasoning:
{ai_reasoning}

Action: {recommended_action}

👉 Claim this lead by replying to this message with "CLAIMED"

First rep to reply "CLAIMED" gets assigned the lead (managed via webhook logic).

Option B: CRM integration via webhooks

For deeper integration with your existing sales tools:

1. Create a webhook pointing to your CRM's API endpoint
2. Use n8n or Zapier to route MoltFlow webhook events to your CRM
3. Automatically create a deal/opportunity in your CRM with:
   • Contact info from MoltFlow
   • Lead score and breakdown
   • Source group and message content
   • Recommended next action

Popular integrations:

• HubSpot: Use HubSpot's Contacts API to create/update contacts with lead score
• Salesforce: Create Lead objects with custom scoring fields
• Pipedrive: Create deals in your pipeline with AI-generated notes

Option C: Email digest for your team

If your team prefers email:

1. Use MoltFlow's Email Reports feature
2. Configure: Daily digest at 9 AM
3. Include: All leads from yesterday with score >= 4
4. Format: Table with score, contact, message preview, recommended action

This works well for teams that prefer batch processing leads rather than real-time alerts.

Step 7: Measure and Optimize Conversion Rates

After running your automated qualification for 2-3 weeks, analyze performance and optimize.

Track key metrics:

In MoltFlow's Analytics page, monitor:

• Lead volume: How many leads detected per day/week
• Score distribution: What % fall into hot/warm/cold tiers
• False positive rate: Hot leads that weren't actually qualified (based on sales feedback)
• False negative rate: Closed deals that were initially scored warm/cold
• Time to first contact: How long between lead detection and sales outreach
• Conversion rate by tier: What % of hot/warm/cold leads convert to customers

Identify scoring gaps:

Meet with your sales team weekly for the first month:

1. Review won deals: What was their initial score? Should it have been higher?
2. Review false positives: What signals misled the AI into over-scoring?
3. Review missed opportunities: Any cold leads that should have been warm/hot?

Refine your OpenClaw scoring prompt:

Based on feedback, update the scoring criteria:

If you have too many false positives (hot leads that aren't qualified):

• Increase the threshold for "purchase intent" from 3 to explicit buying language only
• Add disqualification rules for certain phrases (e.g., "just curious")
• Require at least 2-3 scoring categories to be high, not just one

If you're missing real opportunities (false negatives):

• Lower the hot lead threshold from 7 to 6
• Add more nuanced budget signals (e.g., asking about payment plans indicates budget awareness)
• Consider the group context: premium product groups might have higher baseline intent

A/B test different prompts:

Create two OpenClaw configurations with different scoring prompts and alternate between them for a week. Compare:

• Conversion rates
• Sales team satisfaction
• Time spent on unqualified leads

Keep the prompt that produces the best business outcomes, not necessarily the highest lead volume.

Optimize follow-up sequences:

Track open rates and response rates for each message in your nurture sequence:

• If Day 0 has low response rate, make it more personalized using AI-suggested responses
• If Day 3 has drop-off, try changing the timing to Day 4 or 5
• If Day 7 gets no responses, consider ending the sequence at Day 5

Test different messaging angles:

• Value-focused: "Here's how we help companies like yours..."
• Social proof: "X companies in your industry use our product..."
• Scarcity: "We have limited spots available this month..."

Use MoltFlow's message analytics to compare performance.

Common Qualification Patterns

For B2B SaaS:

Focus on:

• Company size signals (e.g., "50-person team", "enterprise")
• Technology stack mentions (integration needs)
• Decision timeline ("budget approved for Q2")

Adjust scoring:

• Increase weight on "urgency" — SaaS sales cycles favor fast movers
• Add "technical fit" category (0-2) — do they use compatible tools?

For e-commerce:

Focus on:

• Product-specific questions (shows they've researched)
• Shipping location (international vs. domestic affects LTV)
• Bulk order signals ("wholesale", "case of 50")

Adjust scoring:

• Reduce weight on "budget" — most consumers don't discuss budget upfront
• Add "purchase history" (0-2) — returning customers score higher

For professional services:

Focus on:

• Project scope clarity (vague ideas vs. specific needs)
• Authority signals ("I'm the founder", "I make the hiring decisions")
• Reference to previous providers (switching pain indicates high intent)

Adjust scoring:

• Add "project value" category (0-3) — larger projects = higher priority
• Increase "urgency" weight — service engagements often have hard start dates

The best scoring criteria come from analyzing your historical data. What did your best customers say in their first message? Build your AI prompt around those signals.

What's Next

You've now built an automated lead qualification system that runs 24/7, scores every prospect consistently, and routes qualified leads to your sales team in real-time.

Explore these advanced capabilities to expand your system:

• How to Build a WhatsApp AI Support Bot — Add automated customer service to complement your sales automation
• Send Bulk Messages to Qualified Leads — Follow up with batch campaigns targeting specific lead segments
• Connect MoltFlow to n8n — Integrate lead scoring with your CRM and sales tools
• Use MoltFlow's REST API — Build custom lead routing workflows

Ready to automate your sales pipeline? Start your Business plan trial to unlock AI lead scoring and automated routing. View pricing.

Questions? Join our community Discord or check out the MoltFlow documentation for advanced lead management strategies.