Cold Email Infrastructure Mistakes Killing Your Results

Here's What Caught My Eye on LinkedIn This Week

A lot of cold email content on LinkedIn is just people agreeing with each other. "Personalize more." "Be human." "Cold email isn't dead." Great, thanks.

This week was different. There were a few posts with actual specifics worth discussing, some dangerous oversimplifications worth calling out, and at least one story that should terrify anyone running outbound from their company's primary domain. I picked the posts that either gave real tactical value or that I had something meaningful to say about. Here's my breakdown.

The One That Should Scare Every Founder Running Outbound

This is a real story and it happens more than people admit. I've talked to founders who had no idea their entire team's email deliverability was tanked until a customer mentioned they couldn't find emails from them. By that point the damage is already done.

The rule is simple: your primary domain is sacred. Never cold email from it. Set up secondary domains, warm them up properly, and treat them as consumable infrastructure. If you're running any serious volume, you need multiple inboxes across multiple domains. For context, when I'm building a campaign I'll typically run 3-5 domains with 2-3 inboxes each before I start scaling sends. That gives me enough capacity to stay under the threshold that triggers spam flags while still generating real pipeline.

The SPF, DKIM, and DMARC setup is not optional. It's the foundation. If those aren't configured correctly before you send a single email, you're building on sand. Check out my full breakdown of the tech stack that actually supports this at coldemailtechstack2025.

27% Reply Rates and the New Volume Math

The 27% reply rate claim is doing a lot of heavy lifting here without context. Reply rate to what? First email? Entire sequence? Across what ICP? I've seen campaigns hit 40%+ reply rates with hyper-targeted lists of 50 people and campaigns at 2% with blasted lists of 50,000. The number alone means nothing without the denominator.

That said, the volume guidance is solid. 80 emails per day per inbox is a reasonable ceiling if your warming was done correctly. I'd actually push that lower for new domains, closer to 30-40 for the first few weeks. The intent signal point is where the real alpha is right now. Sending to someone who just raised a round, just got promoted, just posted about a relevant pain point - that's where reply rates actually move. Spray and pray to a static list is dead. Signal-based sequencing is where you want to be.

The multichannel layering approach is also genuinely effective. Email plus LinkedIn plus voice note in sequence compounds response rates meaningfully. The key is making each touchpoint feel like a continuation of the same conversation, not a new pitch from a different angle.

The Steve Jobs Email Everyone's Citing

The bonus tip at the bottom is the most valuable part of this post and it's the part most people will skip past. If you cannot explain your offer in two sentences, your problem is not your subject line. It is not your personalization. It is not your follow-up sequence. It is your offer.

I've worked with hundreds of founders on outbound and this is the single most common issue I see. They have a complicated, multi-layered service that requires a 20-minute explanation before it starts to make sense, and then they wonder why cold email isn't converting. The copy is not the problem. The offer architecture is the problem. Before you touch a single email, stress-test your offer. Can you say what you do, who it's for, and what result they get in two sentences? If not, stop and fix that first.

The Steve Jobs framing is fine but it's also a bit of a cheat code for LinkedIn engagement. Attach a famous name to basic principles and watch the likes roll in. The principles themselves are correct though. Short subject lines work. Brevity works. Emotional resonance works. Nothing new here but worth repeating because most people still aren't doing it.

The Cold Email Infrastructure Guide Worth Grabbing

The 23M+ emails number gives this post credibility. Infrastructure is genuinely where most cold email programs fail before they start, and the checklist here is solid. Domain acquisition across multiple registrars is an underrated point. Concentration risk on a single registrar means a single policy change can kill your whole setup. Spread them out.

One thing I'd add to Louis's list: monitor your sending reputation weekly, not monthly. Reputation can erode fast and catching a deliverability drop early lets you pause and recover before the whole domain is burned. Tools like Google Postmaster and MxToolbox are free and take 10 minutes to review. Build that into your weekly ops routine.

For a full breakdown of how I think about the sending stack, including which tools I use at each layer, check out my cold email tech stack guide.

The VC Pitch Email Breakdown That Applies to All Cold Email

The "Dear Investor" opener is a gift to anyone teaching cold email because it's such a perfect illustration of the fundamental mistake. The recipient cannot tell you did any research. You are telling them upfront that this message could have gone to anyone. Why would they respond?

The six-element framework here maps almost perfectly to what I teach for B2B cold email in general. Substitute "personal hook" for a trigger or observation specific to their business. Substitute "traction" for a relevant case study or result you got for a similar client. Substitute "funding status" for a specific, low-friction CTA. The bones are the same regardless of whether you're pitching VCs or selling SaaS.

The most important element in that list is number one: why this person specifically. If you cannot answer that question in one sentence, you are not ready to send the email. I have a client who was sending 200 emails a week and booking almost nothing. We cut the list to 40 per week, added a specific reason why each prospect was being contacted, and reply rates tripled. Volume is not the lever most people think it is. Relevance is the lever.

The "Prove It In The Email" Move That Works

This is a genuinely clever execution. If your value prop is about email deliverability, the fact that your email landed in the inbox is live proof of your product working. You cannot fake that. The prospect experiencing the value prop in the act of reading your email is as good as it gets for a first impression.

The broader principle here is one I call proof of concept in the pitch. Your cold email should demonstrate something about your capability, not just describe it. If you help people write better copy, your email should be the best-written thing in their inbox that day. If you help people with video production, send a 30-second video. If you do web design, make sure your email signature links to something that looks incredible. The medium is part of the message.

Matt is right that the lack of personalization doesn't matter here. That is the exception, not the rule. It works because the relevance is baked into the category. If you sell deliverability software and your email lands in the inbox, every email marketer on the planet is your ICP. But for most offers, you cannot get away with zero personalization. Context and pain have to do more work when you cannot rely on that kind of category shortcut.

The AI Poem Email and the Real Problem With "Personalization at Scale"

The AI poem is a perfect encapsulation of where AI-assisted outreach goes wrong. Everyone has access to the same tools. Everyone is running the same prompts through the same models. The output is different words but the same feel. Prospects have already developed pattern recognition for it. They can sense it even when they cannot name it.

Dmytro's point about Gmail's AI filters catching AI-written emails is worth taking seriously. We are in an arms race where the tools for writing cold email and the tools for filtering cold email are evolving at the same pace. The only durable edge is genuine relevance that cannot be faked at scale. That means doing the research. It means knowing something about the company or the person that the AI cannot surface from a LinkedIn profile and a website homepage.

That said, I'd push back slightly on the "50 accounts, not 5,000" framing as a universal prescription. For high-ticket enterprise deals, absolutely. But for a SaaS product with a $200 monthly price point and a broad ICP, doing deep ABM on 50 accounts is not the right ROI calculation. The answer is not always less volume. The answer is more precise targeting and signal-based triggers so that the people you reach out to at scale are the right people at the right moment. That is different from sending less.

The Cold Call Objection That Becomes a Cold Email Asset

This is excellent. "I already sent you one" is one of the cleanest cold call openers when it is true. It creates continuity, it demonstrates that you did the work, and it completely defuses the objection. The prospect asked you to email them. You did. Now you are following up. That is not aggressive. That is professional.

The sequencing logic here is also right. Email first with a specific trigger, follow-up emails with business cases tied to pain, then call as a natural touchpoint in the sequence rather than a standalone cold outreach. This is how multichannel outbound actually works when done well. Each channel reinforces the others. None of them is trying to do everything alone.

If you want templates that work in this kind of multichannel sequence, I have a set at my cold email follow-up guide that covers the exact structure I use across email, LinkedIn, and phone touchpoints.

The Funding Signal Automation Worth Paying Attention To

This is the right direction. Funding announcements are one of the strongest buying signals in B2B outreach. A company that just raised has budget, has growth targets, and is actively hiring and building. The window for outreach is narrow because every vendor with a pulse is going to see that same TechCrunch article. Getting in within hours instead of days is a real competitive advantage.

The automation stack here is basically what tools like Clay make possible at a more accessible level. You can build signal-based workflows that watch for funding events, job postings, hiring signals, and leadership changes, then enrich the relevant contacts and push them into a sending sequence automatically. This is where the infrastructure investment pays off. The hard part is not the outreach. The hard part is making sure your contact data is clean before it hits the sequence. I use email validation as a mandatory step before anything goes into a campaign.

The Josh Braun "Either Way" Trick

Josh Braun consistently produces some of the most nuanced cold email copy thinking on LinkedIn and this post is a good example. The insight about the end of the email is accurate. Most cold emails build correctly through the opening and body, and then tighten up at the CTA in a way that signals desperation or pressure. The prospect can feel the gear shift.

"Either way" or similar low-pressure closers work because they signal confidence. You do not need this meeting. You are offering something genuinely useful and if they are not interested, that is fine. That posture is more persuasive than aggressive CTAs. It is also, frankly, a better reflection of how good products sell themselves. If your offer requires you to pressure someone into a call, the offer may need more work.

For more on how to structure the CTA section of a cold email so it converts without feeling pushy, see my top five cold email scripts.

The One I'm Calling Out: The "One AI Tool Replaced My Entire SDR Team" Post

I want to separate the useful signal from the noise here. AI writing tools including Claude are genuinely useful for drafting sequences, testing angles, and iterating on copy faster. That part is real.

But "replaced my entire SDR team" for 2-3 qualified calls daily is not a flex. That is a low bar. A single decent SDR should be generating more than 2-3 qualified calls daily if the infrastructure and offer are in place. The automation stack described here is also not something you deploy in an afternoon. 650 N8N workflows is a significant engineering project. The comment bait format of this post is doing more work than the actual content.

The specific claim that deserves scrutiny is "AI Response Manager handles objections automatically." I have built and tested a lot of automated reply handling systems. They work for basic routing, unsubscribes, and simple interest signals. They do not work for nuanced objection handling in a way that actually converts. That still requires a human. Anyone telling you otherwise is selling you something.

The Bottom Line This Week

Three things are consistently true across the best posts this week and consistently ignored in the worst ones.

First, infrastructure is not optional and it is not set-and-forget. Your domain reputation, your DNS configuration, your warming schedule - these require ongoing maintenance. The brands burning their primary domains, landing in spam, and concluding that cold email does not work are almost always skipping this layer entirely.

Second, relevance beats volume. The posts with the highest signal this week all pointed to the same conclusion from different angles. Funding signals. Intent-based targeting. ABM over spray and pray. Personalization that could not have gone to 100 other people. The deliverability problems and the reply rate problems usually have the same root cause: you are sending to people who have no reason to care.

Third, your offer is the foundation. Yurii said it as a bonus tip but it deserves to be the headline. If you cannot explain what you do and who it's for in two sentences, fix that before you write another email. I have helped over 14,000 entrepreneurs generate sales meetings and the single most common reason campaigns fail is not bad subject lines or weak follow-up sequences. It is an offer that requires too much explanation to land in a cold email context.

Start with a clean offer. Build your infrastructure correctly. Target by signal. Write like a human. Follow up. That is the whole game. Everything else is noise.

If you want to see how all of this comes together in actual email copy, my killer cold email templates are the fastest way to get frameworks that have been tested across millions of sends.