Unpopular opinion: your email copy matters way less than your infrastructure
aicoach_amy · 2026-01-07 · 3,010 views
7 years writing cold emails. My unpopular take: infrastructure accounts for 70% of your cold email results. Copy accounts for 30%.
A mediocre email from a properly warmed Google Workspace inbox with correct SPF DKIM DMARC will outperform a brilliantly written email sent from cheap SMTP shared infrastructure. Every single time.
Why? Because the brilliant email never reaches the inbox. It lands in spam. The mediocre email actually gets seen because it was delivered properly.
I have seen this pattern across hundreds of clients. Teams spend weeks perfecting their copy while their emails go to spam. Then they switch to proper infrastructure and suddenly their "old copy" starts working.
Priority order: 1. Infrastructure (dedicated Google Workspace or Outlook inboxes, proper DNS, warmup). 2. Targeting (right ICP, right decision makers). 3. Copy (short, relevant, personalized). Most people have this backwards.
Comments (4)
warmup_wiz · MailReach · 2026-01-20
200 per inbox per day is how you end up in every blacklist known to mankind. The safe limit is 15-20 emails per inbox per day. Google's technical limit is 2,000/day and Outlook is 10,000/day but those are NOT for cold email. You will absolutely torch your domains. Show your client the math — to send 200/day safely you need 10-13 inboxes, not 1.
agencygrind · ScaleOutbound · 2026-01-21
We had this exact conversation with a client last month. They wanted to "maximize" each inbox. We showed them the deliverability data — inboxes sending 50+/day had terrible deliverability under 1% reply rates, inboxes sending 15-20/day had 4%+ reply rates. They shut up pretty quick after that.
solosdr · 2026-01-21
if a client insists on this after you've explained the risks, fire them. seriously. they'll burn your infrastructure and blame you when nothing lands. not worth it
grindgary · 2026-01-22
been there. just do the math for them: 200 emails/day divided by 15 per inbox = ~13 inboxes needed. Outlook inboxes are dirt cheap now so there's literally no reason to overload a single inbox. more inboxes more volume more safely