A post showed up on my LinkedIn feed recently that stopped me mid-scroll. A brand with a 200,000-strong email list had been banned by their ESP. They were sharing the story as a cautionary tale (fair enough), but the lesson they drew from it was so wide of the mark that I had to reread it twice.
Their explanation: they were sending all their email marketing through a single domain. That, apparently, was the problem. The solution, in their view, was to spread sends across multiple domains so that one bad campaign couldn’t contaminate the rest of their infrastructure.
Here’s what actually happened. Three people out of 200,000 were so unhappy with what they were receiving that they didn’t just click the spam button and move on. They went to the trouble of contacting the ESP directly to make a formal complaint. Three people. On a list of 200,000.
And the conclusion was: we need more domain names. 🤷♂️
Why Domain Names Aren’t the Solution
Let’s Talk About What a Formal Complaint Actually Means
Marking an email as spam takes one click. Anyone mildly irritated can do it without thinking. Contacting an ESP to file a complaint is a completely different act. A direct complaint to an abuse desk is considered a more significant signal precisely because it takes effort… more than just clicking a button. When someone goes that far, they’re not mildly annoyed. They feel genuinely wronged, enough to spend time on it.
Three people feeling that way on a list of 200,000 is not a domain architecture problem, but a relationship problem. And no amount of domain splitting fixes that.
For context on just how sensitive the environment has become: the global average spam complaint rate sits around 0.014% of accepted emails, with industry best practice set at 0.02%. Gmail and Yahoo enforce a hard threshold of 0.3%, above which they start rejecting messages outright. Some ESPs issue warnings at 0.2% and suspend accounts above 0.5%. The bar is low. Maintaining a clean list and a respected sender reputation is not optional anymore, but table stakes.
But none of that changes what caused this brand’s problem.
Their domain wasn’t the issue. Their emails were.
Where I Stand on Domain Separation
I want to be precise here, because this isn’t a blanket argument against separating domains.
There are absolutely valid reasons to silo certain email types from others.
Transactional emails (receipts, contracts, password resets, order confirmations) should be sent from a different domain than your marketing and automated campaigns. That separation is sensible and worth doing. If your marketing reputation takes a hit, you don’t want it bleeding into the emails your customers rely on to run their lives. That’s a real risk worth protecting against.
What I’m firmly against is the idea of splitting your marketing email across multiple domains as a hedge against your audience complaining. That logic treats spam complaints as something that happens to you, rather than something caused by you. It treats domain architecture as a substitute for good list management, when the two things aren’t related in any meaningful way.
Several experienced email marketing professionals commented as much under that LinkedIn post, people sending hundreds of millions of emails from a single domain with no deliverability issues whatsoever. The original poster dismissed them as old school. The irony is that building healthy sending habits is the oldest and most reliable playbook in this industry, and it works precisely because it hasn’t been superseded by anything better.
AI Is Closing the Loopholes Anyway
Even if you were convinced that scattering your sends across multiple domains would protect you, the window for that approach is closing fast. Mailbox providers and ESPs are deploying increasingly sophisticated pattern recognition that connects the dots between domains, sending behaviours, content fingerprints, and brand identity, regardless of what name is on the envelope.
You can route the same content through five different domains and the systems on the other end will increasingly recognise it as the same sender. ESPs are now building real-time compliance dashboards and flagging clients who trigger excessive complaints or bounce rates, with account recovery taking weeks when things go wrong.
The infrastructure is getting smarter faster than the workarounds can keep up.
Trying to game deliverability through domain shuffling is the email equivalent of changing your phone number to dodge calls from people you’ve annoyed. Eventually, they recognise your voice.
The Real Problem: Permission Isn’t a Relationship
Here’s where most email marketers go wrong, and it’s a mindset issue more than a technical one. Having someone’s email address (even with full, documented, double opt-in permission) does not mean they want to receive whatever you decide to send, at whatever frequency you decide to send it.
Legal permission and relationship permission are different things.
A subscriber gave you their address for a reason, with a specific expectation. The moment your emails stop matching that expectation, because the content drifted, the frequency increased, the pitch got heavier, or the unsubscribe process became difficult, the trust breaks. And once the trust breaks, you’re not dealing with a permission issue. You’re dealing with a resentment issue.
A spam complaint is one of the strongest negative signals a mailbox provider can receive. Unlike an unsubscribe, which is a controlled exit, a complaint indicates the recipient views your email as unsolicited or abusive.
That distinction matters.
Unsubscribes are feedback.
Spam complaints are accusations.
Three people accusing you, on a list of 200,000, going out of their way to contact your service provider, that tells you something specific about how at least some of your audience experienced your emails. The appropriate response is to understand what happened, not to buy another domain.
What You Should Be Doing Instead
The answer to spam complaints is behavioural, rather than technical.
Here’s where the actual work is:
Make unsubscribing frictionless. If clicking the unsubscribe link is the easiest option, people will use it.
If it’s buried, broken, or leads to a form that takes four steps to complete, people will click spam instead. Always provide a clear, easy-to-find unsubscribe link so recipients can opt out without resorting to spam reports. This is basic. It is also, somehow, still regularly ignored.
Respect the original promise. What did you tell people they were signing up for? Match that.
If your welcome email said “monthly updates” and you’re now sending three times a week, you’ve broken the agreement. Management pressure to send more to sell more is the single most common cause of list deterioration I see across client accounts.
Resist it, or at the very least test it before you commit. Run A/B tests on frequency, not just content, and let the data tell you where the ceiling is before you crash through it.
Use segmentation properly, and use the AI tools that are now available to do it. Modern platforms can identify behavioural patterns across your list and surface meaningful segments without you having to hand-craft every rule.
A subscriber who opened every email in the last 90 days should not receive the same content or the same frequency as someone who hasn’t opened anything in six months.
Different profiles warrant different pressure and different content flows. The tools to do this exist, they’re more accessible than ever, and there’s no excuse for treating a 200,000-person list as a single homogeneous audience.
Clean your list regularly. Email lists decay by about 28% annually, meaning a meaningful portion of any large list goes stale each year.
Inactive subscribers who never engage are not neutral since they drag down your engagement metrics, inflate your list size in ways that look good on a dashboard but mean nothing in practice, and increase the statistical likelihood of complaints.
Remove them, or run a proper re-engagement campaign before you do.
The Number That Should Have Triggered a Different Conversation
Three people out of 200,000 filing formal complaints is a small number in absolute terms. But the instinct to minimise it by blaming domain architecture rather than asking why those three people felt that strongly is where the real mistake was made.
Good email marketing starts with a simple question: do the people on my list actually want what I’m sending them? If the honest answer is “most of them do, but we’ve probably pushed it too far with some segments,” that’s a tractable problem with straightforward solutions.
If the answer is “I’m not sure,” that’s a more serious problem, and adding a second domain won’t help you find the answer.
Sender reputation is built on one thing: whether your audience wants to hear from you.
Domain names, IP addresses, and authentication protocols are all in service of that single underlying truth. None of them replaces it.
Respect your audience. Give them what they signed up for. Make it easy to leave if they want to. Everything else follows from those three things, including your deliverability.
Leave a Reply