Originally shared by Richard Illingworth on Twitter.
Why does cold email advice keep contradicting itself?
If you've spent any time in cold email communities, you've seen it: one expert swears by a specific sending volume, another says the opposite. One tool claims to have cracked Google's filters, another focuses on Microsoft. Everyone has a framework, a playbook, a "proven system."
The uncomfortable reality is that no one has a universal answer — and the people who claim they do are selling something.
This isn't pessimism. It's what years of real testing actually show.
What happens when you run serious deliverability tests?
Controlled experiments across cold email accounts reveal just how unstable the rules are. Same copy, different accounts, same recipients — results vary significantly. Google and Microsoft handle sender reputation, authentication signals, and engagement patterns differently, and those differences shift constantly.
According to Mailchimp's Email Marketing Benchmarks, average email open rates vary by more than 20 percentage points across industries — which hints at how many variables affect whether an email even reaches the inbox, let alone gets opened.
The core problem: by the time you figure out what works, the filters have already moved.
What should you actually do instead?
Accept that cold email is a moving target and build your approach around adaptability:
- Test continuously, not once. What worked last quarter may not work today. Small, ongoing tests beat big periodic audits.
- Track leading indicators. Reply rates, bounce rates, and spam complaints tell you more than open rates alone — especially since open tracking is increasingly unreliable.
- Diversify sending infrastructure. Relying on a single domain or sending account is fragile. Spread risk across multiple domains with proper warm-up protocols — the same setup AI outbound tools lean on to protect sender reputation.
- Watch your authentication fundamentals. SPF, DKIM, and DMARC are non-negotiable. If those aren't clean, no tactic will save you.
- Be skeptical of "guaranteed" systems. If someone is selling a cold email course or tool on the premise that they've cracked the algorithm, ask when they last ran the tests. Deliverability shifts fast.
What's actually consistent in cold email?
A few things hold up regardless of platform changes:
- Relevant, personalized messages to the right people consistently outperform high-volume spray-and-pray campaigns
- Smaller, cleaner lists beat massive, unverified ones
- Engagement signals (replies, not just opens) still matter to inbox placement
- Domain reputation is slow to build and fast to damage
The people doing well in cold email right now aren't the ones who found the secret. They're the ones who test quickly, adapt faster than the filters change, and don't pretend they have answers they don't.
If someone tells you they've solved cold email deliverability permanently, that's your signal to stop listening.