Lose Half Your Leads on Purpose

The Anxiety About "Clean Lists" Is a Luxury Belief

I was on a coaching call recently with an investment banker. Sharp guy. He buys and sells companies for other people, raises capital from private equity funds, deals mostly above the $10 million mark. He'd just set up his cold email infrastructure, he had his LinkedIn profile in decent shape, and he was ready to pour fuel on the engine. He told me he wanted to send a million cold emails a month.

My response: let's do it.

Sending a million emails a month means pulling three million leads. Because you're going to download those leads, verify them, and lose 60% of them. Bad addresses, outdated contacts, people who've changed jobs. You build in the losses from the start.

This is what I call the inventory approach to outbound. Stop treating your lead list like precious inventory you need to protect. Start treating it like a raw material you buy cheap and process at scale. Switching from curated lists to bulk processing sounds like nothing. The results are not.

Why Apollo Leads Are "Trash" - And Why I Still Use Them

Normal Apollo, if you're buying it at retail, runs about two and a half cents a lead. I would never pay that for the quality you get on Apollo. The data has known issues - bounce rates, outdated job titles, contacts who've moved on. Anyone who's run serious volume on Apollo knows this.

You don't have to pay retail. When you scrape Apollo through a tool like ScraperCity's Apollo scraper, you're getting those same leads for around $0.0039 per contact. That's less than half a cent. At that price, it doesn't matter if you lose half of them to verification. It doesn't matter if a third of them are trash. They're cheap enough that waste stops being a problem and becomes a cost of goods - a line item you plan for, not a catastrophe you try to prevent.

Compare that to the agencies agonizing over "list quality" while rationing 1,000 leads like they're hand-selected rare earth minerals. Those guys are optimizing for the wrong thing. They're so scared of waste that they never send enough to figure out what works.

The Three-Million-Lead Breakdown

You want to send one million cold emails a month. Okay. Assume you verify your list before sending - which you should always do, no exceptions. After verification, you're going to lose a significant chunk. Let's say 60%. That means to end up with one million sendable contacts, you need to start with about three million raw leads.

Three million leads a month. Thirty-six million a year. If you want your campaign to run for more than a year or two without burning through your total addressable market, you need a big enough pool in Apollo - we're talking at least a hundred million contacts matching your search criteria. Apollo has north of 275 million contacts in their database, so it's very possible. Make sure your search criteria is that broad - don't add so many filters that you shrink your pool to nothing.

At $0.0039 per lead through the Apollo scraper, three million leads costs you about $11,700 a month. But if your offer closes even a handful of investment banking mandates at deals above ten million dollars, the $11,700 becomes irrelevant.

How I Build Lists - The GPT + Domain Scraper Method

For smaller, more targeted campaigns - say you want to hit a specific niche with higher precision - I use a different workflow that I've been refining. Here's exactly how it goes:

1. Ask GPT to do deep research. Open a new chat, tell it to research-mode and say something like: "Give me the top 100 robotics companies in the United States. List all their domains." It'll go out, scrape the web, and come back with a list. The output is close enough to work with.
2. Feed the domains into a domain email finder. Take that raw list of domains and push them into ScraperCity's email finder. It runs through each domain and pulls back whatever email addresses it can find - usually within a minute or two per batch.
3. Validate the results. Run everything through an email validator before you touch send. Strip out the addresses that will bounce. What you're left with is a verified, qualified list built for a specific niche.

I just did a campaign to robotics companies using this exact method and pulled a 10% reply rate. That happened on a Monday. The list wasn't hand-curated by some data analyst. It was GPT research fed into a domain scraper, cleaned and sent. The message hit because it was written for that specific audience.

When I walked through this on the call, the guy's reaction was basically: where has this been? He'd been overthinking the lead sourcing side when the workflow is pretty mechanical once you've done it a few times. For the full breakdown of how to structure the lead strategy behind these campaigns, check the Best Lead Strategy Guide.

Claude Works for This Too

One thing I mentioned on the call that I want to make sure you catch: Claude outperforms GPT for the deep research list-building step, at least in my experience. Both have a research mode where they'll go out and crawl the web for you. Claude gives you more accurate niche company lists. Try both. See which one gives you better results for your specific vertical.

The workflow is the same regardless of which AI you use. You're just using it as a list compiler - give it a specific brief with the right qualifiers (family-owned, US-based, revenue over X, whatever fits your ICP), let it run for ten or twenty minutes, and you get a raw domain list you can immediately process.

The Conference Angle

I also told this guy about one more channel: trade shows. Skip the broker conferences and the events where other M&A advisors are competing for the same deals. The events where business owners show up - the ones who could be sellers.

The play is simple. You show up without a booth or a sponsorship, talk to everyone, and when someone tells you what they do, you tell them you help companies over $10 million in revenue exit. If they're interested, book a meeting for after the show.

This is especially relevant right now because of the demographic wave happening across American business. The baby boomers are aging out. Gen X is right behind them. A lot of these business owners don't have kids who want to take over the company. They need someone to help them exit - they just haven't been found yet. That's an opportunity hiding in plain sight for anyone willing to show up to the right rooms.

Combined with cold calls - which this guy had no problem with, given his background - you've got a multi-channel motion that compounds. LinkedIn running in the background through Expandi on automation, cold email sending while you sleep, cold calls two hours a day to grind up the deal flow, and conferences to build the warm network. Layer them all and you're coming at prospects from every direction.

Volume Is the Test Itself

One of the things I keep saying that people resist at first: cold email is a volume game now. That doesn't mean the message doesn't matter - I still got a 10% reply rate on that robotics campaign, which proves a good message moves the needle. But the old idea that you can build a list of 200 perfectly curated leads and nurse them toward a response? That's over.

The reason is simple. You don't know what will convert until you test it. You don't know which message, which angle, which offer framing gets replies until it's in front of people. And the only way to find out fast is volume. The guy with unlimited leads finds the winner in a week. The guy rationing a thousand contacts finds it - maybe - in a quarter.

I've been preaching this since I built out my 3C framework - Compliment, Case Study, Call to Action - and it still holds. But even the best framework needs enough reps to optimize. You need to be able to test five different industries, five different messages, and not worry that you're burning through your precious $2,500 worth of leads. When leads cost almost nothing, you can test everything. The Top 5 Cold Email Scripts will help with the message side, but you need enough volume to test against.

Plan for the Losses

Stop trying to build a perfect list. There is no perfect list. Apollo data has known issues and domain scraper results come back incomplete - even a verified list will still bounce on you. Losing leads during verification is how the process works.

When I think about scaling ScraperCity, I think about this exact problem. The agencies that get results with the tool download 50,000 leads, verify them, lose 30,000, and send to the remaining 20,000. They've internalized that the loss rate is just a cost of goods, like a manufacturer who buys raw materials knowing some percentage won't make it through the production line.

When you price your inputs cheaply enough, waste stops being a failure. It becomes a rounding error. You stop being precious and start moving fast. Instead of asking "is this lead good enough to contact?" you ask "how many more can I pull?"

The investment banker on this call got it immediately, because he already thinks this way about deal flow. Not every deal he looks at will close. Most won't. But he keeps his pipeline full because he knows his conversion rate and he works backward from there. Cold email works exactly the same way. Know your reply rate and your meeting rate, know your close rate - then build a list big enough to hit your number even after the losses.

The Practical Checklist Before You Scale

Before you go chase three million leads a month, get these fundamentals locked in:

• Your Apollo search criteria has to be big enough. If you're sending a million emails a month to the same pool, you'll exhaust it. Check the total addressable market in Apollo before you commit to a sending volume. You need to be able to sustain this for years, not weeks.
• Verify before you send. Every single time. No exceptions. Sending to unverified lists destroys your domain reputation and all the volume in the world won't save you from that.
• Have a clear, specific offer. "I help companies over $10 million exit for great multiples" - that's specific enough to get a reply. "I do marketing consulting and growth strategy" is not. Volume amplifies whatever you're saying. If what you're saying is vague, volume just means more people ignoring you faster.
• Set up your sending infrastructure before you scale. Multiple inboxes, proper warm-up, delivery tools like Smartlead or Instantly - get all of this dialed in at low volume first, then ramp.
• Have a CRM ready for replies. At this volume, managing responses will bury you. Close is what I use - simple pipeline, fast. Whatever you use, have it ready before your inbox gets flooded.

The Bottleneck Isn't Leads

The investment banker told me at the end of the call: "I'm going to send a million emails a month, dude." My answer was: go for it. Make sure your search is big enough. Make sure the criteria is right. And be ready for what happens when the replies start coming in, because at that volume, your hardest problem becomes replying to everyone who's interested.

That's a good problem to have. And the only way to get there is to stop treating your lead list like a precious resource you have to protect - and start treating it like cheap raw material you process at scale.

Download more than you need, verify them, and send whatever survives.

If you want the full framework for how to structure the outreach system behind this - the scripts, the follow-up sequences, all of it - grab the Cold Email Manifesto. That's where I lay out the complete methodology. And if you want to work through your specific campaign with a coach who can look at your offer and your list together, check out Galadon Gold - that's where we do this work live.