Use Case: E-commerce & Lead Gen
price monitoring · competitor scraping · lead qualification · contact enrichment
What you'll build
An agent that monitors competitor pricing daily, alerts when prices drop below yours, scrapes product data into a spreadsheet, and qualifies new leads by enriching them with company size, tech stack, and LinkedIn data.
Scraping ethics: Only scrape public pages where no ToS explicitly prohibits automated access. Respect robots.txt. Add a delay between requests (the browser-control skill does this automatically). For high-volume scraping, use a dedicated scraping API like Apify — see the Apify integration guide.
Jump to section
Skills you need
| Skill | What it unlocks | Required? |
|---|---|---|
| browser-control | Scrape dynamic pages (React/SPA sites) that a simple HTTP fetch can't read | Core |
| http-request | Pull pricing or product data from JSON APIs — faster than browser for API-based sites | Core |
| structured-extractor | Extract specific fields (price, SKU, availability) from page HTML into JSON | Core |
| file-writer | Save scraped data to CSV for import into Google Sheets or Excel | Recommended |
| web-search | Find new competitor pages, product launches, or press releases | Recommended |
| notion-writer | Push qualified leads directly into a Notion CRM database | Optional |
SOUL.md template
# E-commerce & Lead Gen Agent You monitor competitors, track prices, and qualify leads. ## Identity - Name: SalesBot - Role: Competitive intelligence and lead qualification ## Price monitoring rules - Check competitor pricing pages listed below - Extract: product name, price, plan name, any promotional pricing - Compare against our current prices (listed below) - Alert if a competitor's equivalent plan is more than 10% cheaper than ours ## Our current prices - Starter: $29/mo - Pro: $79/mo - Enterprise: $199/mo ## Competitor pages to monitor - Competitor A: https://competitor-a.com/pricing - Competitor B: https://competitor-b.com/pricing ## Lead qualification criteria A lead is QUALIFIED if: - Company size: 10–500 employees - Industry: SaaS, fintech, or e-commerce - Has a technical role (CTO, VP Eng, Lead Dev) in the contact A lead is UNQUALIFIED if: - Company size < 10 employees (too small) - Industry: government, non-profit, education ## Output format for leads For each lead: Name | Company | Size | Industry | Role | Score (1-10) | Recommended action
Price monitoring setup
{
"skills": [
"official-browser-control",
"official-http-request",
"official-structured-extractor",
"official-file-writer"
],
"model": "claude-sonnet-4-5",
"soulPath": "./SOUL.md",
"browser": {
"headless": true
},
"channel": {
"type": "telegram",
"token": "YOUR_BOT_TOKEN",
"chatId": "YOUR_CHAT_ID"
}
}Check competitor pricing now
"Visit all competitor pricing pages in my SOUL.md. Extract plan names and prices into a markdown table. Flag any plan cheaper than ours by more than 10%."
Scrape product listings
"Scrape the first 3 pages of https://competitor.com/products — extract product name, price, rating, and review count. Save as ~/data/competitor-products.csv"
Find new competitors
"Search for "project management tool for developers" on Google. List any tools you haven't seen before that appear in the top 20 results."
Lead qualification workflow
Pass a list of leads (from a CSV export, CRM webhook, or paste) and the agent scores and enriches each one:
Qualify from CSV
"Read ~/leads/march-signups.csv (columns: name, email, company). For each lead, find the company's LinkedIn page, check employee count and industry. Score each 1-10 using my SOUL.md criteria. Save the scored list as ~/leads/march-signups-scored.csv"
Qualify a single lead
"Qualify this lead: Sarah Chen, CTO at TechFlow (techflow.io). Find their company size, tech stack from their job postings, and any recent funding. Score and recommend next action."
Enrich existing CRM data
"Read ~/crm-export.csv — it has company names but no employee count or industry. Enrich each row by looking up the company on LinkedIn and Crunchbase. Save the enriched file."
Scheduled competitor reports
{
"crons": [
{
"name": "Daily competitor price check",
"schedule": "0 8 * * 1-5",
"task": "Check all competitor pricing pages from SOUL.md. If any competitor has dropped their price more than 10% below ours, send an alert to Telegram with the details. Otherwise send a brief 'no changes' summary."
},
{
"name": "Weekly competitive report",
"schedule": "0 9 * * 1",
"task": "Check all competitor pricing pages and their blog/changelog for new product announcements from the last 7 days. Post a competitive intel summary to Telegram."
}
]
}Common issues
⚠ Competitor pricing page is behind a login
If the pricing page requires login, the agent can't scrape it. Use a tool like Google Cache or archive.org for last known prices, or check the pricing via their public API if one exists.
⚠ Scraper gets blocked after a few pages
Add a rate limit note to your SOUL.md: "Wait 3 seconds between page loads." For high-volume scraping, switch to the Apify integration which handles rotating proxies automatically.
⚠ Lead enrichment returns outdated company data
LinkedIn data accessed via browser can be 30–90 days stale. Specify in your prompt: "Check the job postings page for current company size if the main page doesn't show it." Or use a dedicated enrichment API.
⚠ CSV output has inconsistent formatting
Tell the agent exactly which columns you need: "Output CSV with columns: name, company, size, industry, role, score, action — in that order, no extra columns."
Did this guide solve your problem?