AI Dealer Intelligence System: Score, Filter & Auto-Message Dealers

A complete AI dealer intelligence pipeline: from raw marketplace data to scored leads with automated WhatsApp outreach · Photo: Pexels

The Problem: Sales Teams Flying Blind on Dealer Data

Picture this: your sales team has just pulled a list of 5,000 potential dealers from a marketplace. Hardware shops, fabricators, distributors, construction suppliers - all mixed together with no signal on which ones are actually worth calling. They start dialing randomly. Three hours later, they've reached 40 businesses, 30 of which were completely irrelevant.

This is the default state for most B2B manufacturers and distributors. Dealer acquisition is manual, slow, and expensive - not because the data doesn't exist, but because no one has built a system to prioritize it intelligently.

The AI dealer intelligence pipeline solves this entirely. Every lead gets a score. Every score maps to an action. Every action triggers automatically - without a human in the loop.

What This System Can Actually Do (Global Scope)

This is not a tool limited to one country or one marketplace. The pipeline is designed to work with any publicly listed dealer data source, anywhere in the world.

Supported Data Sources

Scale: No Upper Limit

• Firecrawl can process hundreds of thousands of pages per crawl job
• Google Sheets handles up to ~500,000 rows comfortably
• For larger datasets, swap Sheets for BigQuery or Supabase - same scoring logic applies
• Twilio WhatsApp scales to millions of messages per month on a business account

System Architecture Overview

Five components, all free or near-zero cost at the core level:

The AI Dealer Scoring Engine (1–10 Logic)

Every dealer gets a score from 1 to 10 based on three weighted signals: business type fit, location tier, and verification status. The score determines the bucket - and the bucket determines the automated action.

Base Score by Business Type

Customize the categories and weights to match your product. The example below is for a building materials manufacturer:

Score Modifiers

The Four Dealer Buckets

Step 1: Scrape Dealer Data from Any Marketplace

Firecrawl is a web crawling API that returns structured data from any website. Point it at your marketplace search results and it returns clean JSON or CSV - no manual copy-paste needed.

// Example search URL patterns (replace with your marketplace + product)
// IndiaMart (India)
https://www.indiamart.com/search.mp?ss=YOUR+PRODUCT+KEYWORD

// Thomasnet (USA)
https://www.thomasnet.com/search/?searchterm=YOUR+PRODUCT

// Alibaba (Global)
https://www.alibaba.com/trade/search?SearchText=YOUR+PRODUCT

// Google Maps (any country)
https://www.google.com/maps/search/dealer+YOUR+PRODUCT+CITY
URL patterns

Required CSV Column Structure

Your scraped output must have at minimum these columns for the scoring engine to work:

Company, City, State/Region, Phone, Source URL, Verified (GST/VAT/Tax)

// Example rows (fictional companies for illustration):
"RoofTech Supplies Ltd",   "Example City", "Region A", "98XXXXXXXX", "https://...", "Verified"
"AlphaFab Works",           "Example City", "Region B", "70XXXXXXXX", "https://...", "Not Verified"
"Metro Hardware Depot",     "Metro City",   "Region A", "91XXXXXXXX", "https://...", "Verified"
CSV

Step 2: Score & Filter with Google Apps Script

The scoring logic runs inside Google Apps Script - a free JavaScript environment attached to your Google Sheet. Here's the complete, adaptable scoring function:

function scoreDealer(businessType, city, verified, competitors) {
  let score = 5.0; // default baseline for unknown types

  const typeMap = {
    'roofing':     9.5,  'sheet':        9.5,
    'aluminium':   9.0,  'glazing':      9.0,
    'hardware':    8.5,  'timber':       8.5,
    'fabrication': 8.5,  'steel':        8.5,
    'plastic':     8.0,  'polymer':      8.0,
    'engineering': 7.0,  'construction': 7.0
  };

  const lowerType = businessType.toLowerCase();
  for (const [keyword, baseScore] of Object.entries(typeMap)) {
    if (lowerType.includes(keyword)) { score = baseScore; break; }
  }

  // Auto-remove competitors
  const lowerComp = competitors.map(c => c.toLowerCase());
  if (lowerComp.some(c => lowerType.includes(c))) {
    return { score: 0, bucket: 'IGNORE', flag: 'COMPETITOR' };
  }

  if (verified === 'Verified') score += 0.5;

  const tier1Cities = ['metro', 'capital', 'port'];
  const lowerCity = city.toLowerCase();
  if (tier1Cities.some(c => lowerCity.includes(c))) score += 0.3;

  score = Math.min(score, 10);

  const bucket =
    score >= 9.0 ? 'HOT AUTO LEAD' :
    score >= 7.5 ? 'VERIFY FIRST'  :
    score >= 5.0 ? 'NURTURE'        : 'IGNORE';

  return { score, bucket, flag: 'OK' };
}
Google Apps Script

Step 3: Build the 5-Tab Live Dashboard

Organize the Google Sheet into five purpose-built tabs, each serving a different function in your sales workflow:

Step 4: Automate WhatsApp Outreach with Twilio

Once dealers are scored and bucketed, the system fires personalized WhatsApp messages via Twilio's WhatsApp API. Two flows run automatically: outreach to dealers, and internal alerts to your sales reps.

Dealer Outreach Message Function

function sendDealerOutreach(phone, companyName, city, productCategory) {
  const ACCOUNT_SID = 'YOUR_TWILIO_ACCOUNT_SID';
  const AUTH_TOKEN  = 'YOUR_TWILIO_AUTH_TOKEN';
  const FROM_NUMBER = 'whatsapp:+1XXXXXXXXXX';

  const messageBody = [
    `Hello, we are reaching out to ${companyName} in ${city}.`,
    `We manufacture ${productCategory} and are looking for authorized`,
    `dealers in your area. Interested in a partnership?`,
    `Reply YES for product details and pricing.`
  ].join(' ');

  const endpoint = `https://api.twilio.com/2010-04-01/Accounts/${ACCOUNT_SID}/Messages.json`;
  const credentials = Utilities.base64Encode(`${ACCOUNT_SID}:${AUTH_TOKEN}`);

  UrlFetchApp.fetch(endpoint, {
    method:  'post',
    headers: { Authorization: `Basic ${credentials}` },
    payload: { From: FROM_NUMBER, To: `whatsapp:+${phone}`, Body: messageBody }
  });
}
Google Apps Script

Step 5: Deploy, Trigger & Run on Autopilot

Open your Google Sheet → Extensions → Apps Script. Paste in all functions (scoring, bucketing, formatting, Twilio). Save and run once to grant permissions. Then set a master function:

function runDealerIntelligence() {
  clearAllSheets();
  const raw    = loadDealersFromCSV();
  const scored = scoreDealers(raw);
  writeToAllSheets(scored);
  applyBucketFormatting();
  sendHotLeadAlerts(scored);
}
Google Apps Script

In Apps Script → Triggers, add a time-driven trigger: run runDealerIntelligence every morning at 8 AM. Your team wakes up to a fully scored, color-coded, WhatsApp-ready lead dashboard - every single day.

Sample Output: Scored Dealer Table

Below is an example output table using fictional company names to illustrate what the scored dashboard looks like in practice.

All company names above are fictional examples only. Competitor dealers are auto-flagged and removed before this table is generated.

Scaling Beyond Google Sheets

The scoring logic - the JavaScript functions - remains identical at every scale level. Only the data storage and trigger layer changes.

What's Next: Make.com, AI Personalization & CRM Sync

This guide is the foundation. Each of these layers can be added incrementally - start with the sheet, add Make.com when your volume grows, and layer in AI personalization when your messaging needs to feel more human.

Whether you are building your first AI dealer intelligence system or scaling an existing B2B sales automation pipeline, the approach in this guide gives you a complete, repeatable foundation. From Google Apps Script lead scoring to WhatsApp automation and beyond, every component is designed to run without ongoing manual effort - so your sales team focuses on closing, not chasing.