Your Lead Source and Infrastructure Both Stink

I Watched Two Cold Email Videos So You Don't Have To

I watch a lot of cold email content. Most of it is recycled advice from people who sent their first campaign two or three years ago and now sell courses about it. But occasionally something good shows up on my feed.

This week, two videos stood out. They're both attacking the same invisible problem from different angles: the default cold email playbook is broken, and it's not obvious where it's broken.

One video argues you're emailing the wrong people. The other argues your emails aren't even landing. Both are right, and the combination is more powerful than either video alone.

Let me break them down.

Video 1: Why Your Reply Rates Are Stuck at 2%

The premise of this video is simple and correct: the same B2B databases feed the same filters, producing the same overcontacted decision makers. And those people have built up so much immunity to cold email that no amount of clever copywriting is going to break through.

The quote that stood out to me: "I'm so hardened by it that there's no cold email in the world that you could send me that's going to get a 10% reply rate. No way on earth."

I feel this in my bones. I get hundreds of cold emails a day. I have filters set up to route most of them into a folder I check once a week out of professional curiosity. And I'm not unique. I talk to VPs of Marketing, agency owners, founders with a LinkedIn profile and 50+ employees - they all describe the same inbox overload. They're drowning in cold email from people who all found them on the same platform, using the same title filters, in the same industry buckets.

The solution presented is to stop using B2B databases like Apollo and ZoomInfo as your primary lead source. Instead, go after people who aren't being cold emailed 100 times a day.

The Three Alternative Lead Sources

The video walks through three tiers of alternative data, and I want to give my honest take on each one.

Tier 1: Google Maps / Local Business Leads. The argument is that local business owners, like restaurant owners or junk removal companies, barely get cold emails compared to the typical B2B prospect. So when you do email them, your reply rate jumps from 2% to 4% or higher just because of reduced competition in their inbox.

This is accurate. This plays out with clients at every stage. Local leads respond at significantly higher rates than the standard Apollo export because they simply aren't used to receiving outbound. A restaurant owner in Austin getting a well-written email about a service that could help them is going to read it. An agency owner in New York getting their 47th "we do SEO" email today is not.

The challenge with local leads is that many of them don't have individual email addresses publicly available. You're often working with role-based emails like info@ or contact@, which can still work for smaller operations where the owner checks everything. But for bigger franchises, those go to a support team. If you're going after local leads, you need a way to find the actual owner's contact info. Tools like ScraperCity's Google Maps tool or email finder can help you find those owner emails.

Tier 2: Creator and Influencer Leads. This is where the video gets interesting. The creator economy is massive, and creators rarely show up in traditional B2B databases. You can't go into Apollo, type "YouTuber," and expect to find the right people. Their emails are different. Their buying patterns are different. And critically, they're not getting hammered with the same outbound as a SaaS founder.

I agree with this in principle, but with a caveat. Creator leads work incredibly well if you have an offer that's relevant to creators. Content production services, sponsorship platforms, merch solutions, financial planning for self-employed people. But don't force it. If you sell enterprise cybersecurity software, emailing YouTubers is not going to save your campaign. The lead source has to match the offer.

Tier 3: Instagram Database Leads. The video mentions an Instagram-based lead database as the highest-performing tier, and this is the one that drove the 16% reply rate case study for a peptide company.

The logic is sound. Instagram business accounts, especially in health, wellness, ecommerce, and DTC niches, represent a pool of business owners who are almost never targeted through traditional cold email channels. These people have their guard down because they get far fewer outbound emails than B2B decision makers.

Where I Agree and Where I Push Back

The core thesis is right. Lead source beats copy. I've been saying this for a long time, and it's one of the things I cover in my top cold email scripts. The best script in the world sent to the wrong list will underperform a mediocre script sent to a perfectly targeted, undercontacted list. Every time.

We've worked with over 14,000 clients across every niche. Same story every time. When someone comes to us stuck at 1-2% reply rates, the first thing we look at is their list. Where did the leads come from? What filters did they use? How many other people are probably emailing the same contacts?

Where I push back is on the framing that B2B databases are useless. They're not. Apollo and ZoomInfo and similar platforms still have a role, especially for enterprise outbound where you need to reach specific titles at specific companies. Lazy filtering is the problem. They slap on three filters, export 10,000 contacts, and blast the same template to all of them.

If you're going to use a B2B database, you need to layer in additional qualification. Job postings, recent funding, tech stack signals, hiring patterns. Use Clay to enrich your leads with context that tells you whether this person has the problem you solve right now. That's what separates a 2% campaign from a 5% campaign even within B2B data.

But if you want to get above 5%? The video is right. You need to find undercontacted segments.

Video 2: The 5-Level Inbox Testing System

This video takes a completely different angle. Instead of asking "who are you emailing," it asks "are your emails even getting there?" And the answer is almost always: you have no idea.

The framework presented is a five-level system for testing inbox health, and it's one of the more rigorous approaches laid out publicly. Let me walk through each level and add my take.

Level 1: Warm-Up Reputation Scores

This is usually the only check anyone runs. You look at your warm-up reputation score in Instantly or SmartLead, see a number above 90%, and assume everything is fine.

The video's stance: anything below 99-100%, pull the inbox. I agree with being aggressive here, but I'd add context. Warm-up reputation scores are a lagging indicator at best and misleading at worst. They tell you how your inbox performs against other warm-up inboxes, not against real-world recipients. It's like testing your car's top speed on a racetrack and assuming it'll perform the same in city traffic.

Use it as a first filter, not a final answer.

Level 2: Seed Inbox Testing

This involves sending test emails to a set of inboxes controlled by your sending platform to see where they land (inbox vs. spam). Better than Level 1 because you're testing deliverability, not just warm-up performance.

The limitation, as correctly pointed out in the video, is that seed inboxes have never marked anything as spam before. They lack any history or behavioral patterns. So passing a seed inbox test doesn't mean you'll land in the inbox of a real prospect whose email provider has learned their spam preferences over time.

Do it anyway. Just don't stop here.

Level 3: Reply Rate by Domain

Now we're getting into the good stuff. The system described runs an automated check every Friday that looks at each domain and inbox: how many emails sent, how many replies received over a one and two week window. Any inbox that has sent at least 100 emails with less than 1% reply rate gets cut.

This is exactly right and something I recommend to everyone running cold email at scale. Your reply rate per domain is the single most honest signal you have about inbox health. If one domain is getting 3% replies and another is getting 0.2%, the second one is probably going to spam regardless of what the warm-up score says.

Automate this. If you're managing 50+ inboxes, you cannot do this manually every week. Set up the automation. Use Clay, use Claude, use whatever. Just make sure it runs consistently. I talk about the full tech stack for this in my cold email tech stack guide.

Level 4: Auto-Response Inbox Testing

This is clever.

The idea: collect all the automated "this email is no longer monitored" responses you get from campaigns. Build a list of these auto-reply inboxes (they mentioned having around 10,000 of them). Then use those inboxes as your own seed testing list. When you send to them, you should get an automated reply back. If an inbox can't even trigger an auto-response from a mailbox that auto-responds to everything, that inbox is dead.

The threshold mentioned is 80% auto-reply rate expected, with anything below 60% being a kill signal.

I love this approach because it uses inboxes with spam histories instead of the pristine seed inboxes from Level 2. It's a much more honest test of whether your emails are reaching anyone.

The effort to set this up is non-trivial, but once you have the list built, it runs itself. Start saving every auto-response you get from every campaign. Within a few months, you'll have a substantial testing list.

Level 5: Positive Behavior Campaigns

The final level is using your sending inboxes for high-response activities beyond cold outreach. The examples given are candidate sourcing emails and partnership outreach, both of which naturally generate high reply rates.

The theory: by generating positive engagement on an inbox (not just warm-up engagement, which Google almost certainly recognizes as artificial), you extend the inbox's useful life and build a stronger sender reputation.

The diagnostic benefit is also valuable. If your partnership campaign that normally gets 15% replies suddenly drops to 3% on a specific inbox, you know that inbox has a deliverability problem before it tanks your main campaign.

This is smart and something I hadn't fully systematized in my own operations. We've done something similar intuitively, using certain inboxes for both outbound and legitimate business correspondence, but formalizing it as a diagnostic layer is a level up.

The Insurance Inbox Strategy

Beyond the five levels, the video introduces the concept of "insurance inboxes," keeping 50% extra sending capacity warmed up and ready to deploy when active inboxes die.

So if you need 100 inboxes to send 3,000 emails per day (30 per inbox), you keep another 50 inboxes warming in reserve. When active inboxes fail your five-level tests, you swap in insurance inboxes with zero downtime.

This is cold email infrastructure done properly. It's tempting to treat inboxes as permanent assets. Inboxes are consumable resources with a finite lifespan. Plan accordingly.

Hypertide offers free domain warming for reserve inboxes - useful if you're running at scale. Paying for inboxes that sit idle just to stay warm is one of the hidden costs of cold email.

The Domain Forwarding Question

One more point from this video. There's a discussion about whether forwarding your cold email domains (like go.yourbrand.com) to your main domain could create blacklisting issues.

The practical answer given is that Google and Outlook aren't doing this yet, and the video makes a great point about why: if forwarding domains could tank a main domain's reputation, bad actors would just buy domains, forward them to competitors, and send spam to destroy their email reputation. The attack surface would be too large for Google to act on this.

Some third-party blacklist providers are catching on, but as noted, Google recently announced they don't rely on external blacklists like Sorbs for their spam decisions. So for now, forwarding is still fine. But spinning up microsites on Railway or similar platforms as an alternative is cheap insurance if you want to be cautious.

The Throughline: Default Cold Email Is a Losing Game

Both of these videos are attacking the same core problem from opposite ends.

Video one says: you're emailing the wrong people because you're fishing in an overcrowded pond. Switch your lead source and your reply rates will multiply without changing anything else about your campaign.

Video two says: even if you're emailing the right people, you might not be reaching them. And the standard tools for checking deliverability only show you the surface. You need five layers of testing to really know what's happening.

Both problems compound. Bad lead source plus unreliable infrastructure equals zero results. Fix one and you improve. Fix both and you're operating at a completely different level than 95% of the people sending cold email today.

Thousands of client engagements confirm this. The people who get consistently high reply rates aren't doing anything magical with their copy. They obsess over lead sourcing. They obsess over whether their emails land.

If you're struggling with reply rates right now, don't start by rewriting your email for the 15th time. Start by asking two questions:

• Where did my leads come from, and how many other people are probably emailing the same contacts?
• Do I know which of my inboxes are landing in primary inbox versus spam?

If you can't answer both of those with confidence, fix that before you touch your subject line.

What to Implement This Week

From Video 1, take one segment of your current campaign and rebuild the list using a non-B2B-database source. Google Maps leads for local businesses, creator leads if your offer fits, or even industry-specific directories that get overlooked. Run it side by side with your existing list using the same copy. Let the data tell you whether lead source is your bottleneck.

From Video 2, set up the Level 3 automation at minimum. Every Friday, pull reply rates by domain and inbox. Cut anything below 1% reply rate after 100+ sends. If you're using Instantly or SmartLead, this can be automated in under an hour. Then start collecting auto-response emails for your Level 4 testing list. It'll take a few weeks to build up, but it's the most underused deliverability signal in cold email right now.

And if you need scripts to run against your new lead sources, grab my latest email scripts pack. They're built for the campaigns I'm running right now.

Stop fighting over the same overcontacted inboxes. Stop trusting warm-up scores as your only health check. Fix the foundation, and the reply rates follow.