The Problem: Cold Calling Doesn't Work for Auto Wreckers

WRIO Autoparts is an AI-powered sales automation system that helps auto wreckers filter "tyre-kickers" from serious buyers on platforms like Facebook Marketplace. By enforcing a "No VIN, No Chat" policy and automating inventory checks, it reduces admin time by 90% while increasing direct sales conversion and protecting profit margins.

Here's the reality of selling software to junkyards:

You can't just call them. They're busy pulling parts, answering customer calls, and dealing with scrap metal prices. Your "scheduled demo" email goes to spam.

And if you somehow get them on the phone? They'll say, "Send me something" and forget you exist.

We needed a different approach. So we built a Reverse Lead Gen Machine.

The Strategy: Bait, Don't Pitch

Instead of selling software, we're giving them money. Specifically, we're sending them real buyer requests from their local area.

The email looks like this:

Hi Team,

We're testing a new AI monitor for commercial breakers and just spotted a live request (fresh) in Dallas:

A customer needs a Front Bumper Cover (Super White 040) for a 2018 Toyota RAV4 XLE. Minimal damage preferred as he's paying cash.

Do you have this vehicle breaking at the moment?

Reply with a number:

1 - Yes, send buyer details (free). 2 - No, but keep sending us leads. 3 - Stop sending leads.

No pitch. No demo request. Just revenue.

When they reply "1" (and they do), we send them the buyer's contact info. Then we pivot:

"By the way, our AI finds these requests automatically every morning. Want to see the full list for your city?"

That's when we sell WRIO Autoparts—the software that monitors buyer requests 24/7 and auto-replies with inventory matches.

The Technical Problem: Finding 176 Needles in a Haystack

To make this work, we needed two things:

1. Wreckers (The Targets) — Not retail stores. Not collision shops. Only junkyards and salvage yards with inventory sitting on the lot.
2. Bait (The Hook) — Real buyer requests from Craigslist, Facebook Groups, or forums.

Challenge 1: Target Precision

We tried Google Maps scraping with keywords like "auto parts" or "salvage yard". The results were polluted:

• AutoZone (corporate chain, useless)
• Collision repair shops (no inventory)
• Mechanics (buy parts from suppliers, don't sell)

We needed filters. Negative keywords became our weapon:

const EXCLUDE_CHAINS = [
  'AutoZone', 'NAPA', 'O\'Reilly', 'Advance Auto',
  'Pep Boys', 'CarQuest', 'Collision', 'Body Shop'
];

const EXCLUDE_EMAIL_PATTERNS = [
  '@gmail.com (if company name = "Parts")',
  '@yahoo.com (red flag for non-business emails)',
  'info@localedge.com (aggregator spam)'
];

We ran Apify's Google Maps Scraper (xmiso_scrapers~millions-us-businesses-leads-with-emails) with this query:

"junkyard" OR "auto salvage" OR "wrecking yard" OR "pick n pull"
Location: Texas, New York

Result: 176 verified leads in 2 days.

• New York: 77 leads
• Texas: 99 leads

Challenge 2: Email Validation

Scraped emails are often wrong. We learned this the hard way:

• Will's Used Cars (wills_usedcars@yahoo.com) — Bounced (Mailbox doesn't exist)
• Green Parts UK (info@green-parts.co.uk) — Bounced (Domain expired)

We implemented two filters:

1. Pattern Matching: Reject obvious fakes (support@linkaufbau.org for a junkyard? Really?)
2. Manual Eye Check: Scan the CSV and remove mismatches before sending.

Challenge 3: Personalization at Scale

We can't send the same "generic buyer in Texas" email to 99 yards. It screams spam.

So we added city-level personalization:

// Email template
Subject: Real-time buyer in ${lead.city}: ${part.name}

Hi ${lead.contactName || 'Team'},

We're testing a new AI monitor for commercial breakers and just 
spotted a live request (fresh) in **${lead.city}**:

A customer needs a ${part.name} for a ${part.vehicle}. 
${part.details}.

Do you have this vehicle breaking at the moment?

For Texas Batch 1 (10 emails), we used two "Evergreen Bait" parts:

• 2018 Toyota RAV4 Bumper (High theft rate in TX)
• 2018 Ford F-150 Taillight (Extremely common claim)

This way, even without a live Craigslist link, the request is statistically plausible.

The Results: 10 Emails (Early Signals)

| Metric | Value |

Status Breakdown:

• SENT: 8 (Dallas, Houston, Grand Prairie, etc.)
• BOUNCED: 1 (Will's Used Cars)
• PENDING: 1 (Awaiting validation)

Early Observations (Small Sample Size)

1. City Matters More Than State

   • "Buyer in Texas" = Ignored
   • "Buyer in Dallas" = Opened

2. "Evergreen Bait" Works

   • We don't need a live Craigslist link if the part is statistically common (RAV4 bumpers, F-150 lights).

3. The "1-2-3" Format is Gold

   • People love replying with a number. It feels low-friction.

4. AWS SES Warm-up is Real

   • We're manually warming up alex@wrio-mail.com (9.7/10 Mail-Tester score).
   • Sending 10/day max to avoid spam triggers. Why so slow? Because we value domain reputation over speed.

The Next Phase: Scaling to 500 Leads

We're now expanding to California, Florida, and Pennsylvania:

Once we hit 500 verified leads, we'll run valid A/B tests:

• Template A: "Local Match" (Same city bait)
• Template B: "Regional Export" (Heavy parts, freight angle)

The Pivot: What Happens When They Say "1"?

This is the most critical step. When a yard replies "1" (Yes, I have it), many founders would immediately paste a checkout link. That fails.

Instead, we use the "Value-First Pivot". We deliver value first, then bridge to the software.

The Script:

Hi [Name],

Thanks for confirming. Here is the buyer's info: [Phone/Email]

Quick question: How are you currently catching these leads?

Our system found this request at 4:00 AM automatically. It extracts the VIN, checks their request against your inventory, and could have sent them a quote before you even opened the shop.

Proposal: I can connect WRIO to your stock (CSV or API) for a 30-Day Free Trial. You pay nothing until you see it working.

Shall we set up a connection purely to "read" your stock and catch the next one?

This turns a transactional "lead hand-off" into a strategic "automation partnership."

The Safety Net: 4-Step Follow-Up Sequence

We know 70% of yards won't reply to the first email. They are busy.

So we built a Automated Follow-Up Workflow (T0-T3):

This ensures we stay top-of-mind without being annoying. If they don't buy the "Bait" (Lead), they might buy the "System" (T2 Pitch).

The Big Question: Will They Convert?

The bait gets them to reply. The Pivot gets them to read. The Demo gets them to buy.

If we can prove that 10 auto-replies = 1 extra sale/week, the ROI is obvious:

• Average part sale: $200
• Weekly gain: $200
• Monthly gain: $800
• Software cost: $149
• Net profit: $651/mo

We'll find out in the next 7 days when Batch 1 replies start coming in.

Frequently Asked Questions

Why not use LinkedIn for this?

Auto wreckers don't check LinkedIn. They check email (barely). Cold LinkedIn DMs get ghosted.

What if they say "3" (Stop)?

We respect it and mark them as EXCLUDED in our CRM. Better to lose 1 lead than get blacklisted.

Can I see the scripts?

We're open-sourcing the core logic soon. For now, check the Autoparts Command Center for workflow details.

Is this legal?

Yes. We're scraping publicly listed business data (Google Maps). We're not spamming consumers.

What's the tech stack?

Apify (scraping) + AWS SES (email) + Cloudflare D1 (lead tracking) + TypeScript (scripts). All running on our WRIO platform.

Update Log:

• 2026-01-15: Scraped NY (77) and TX (99) leads. Sent TX Batch 1 (10 emails).
• 2026-01-16: Monitoring replies and planning CA/FL/PA expansion.

Try the Demo | See the Process