Nothing else you build in outbound matters much once the sending infrastructure goes quiet. I learned that the hard way before I learned it properly.
What broke
At UpHire I owned the whole sending layer: domains, SPF, DKIM and DMARC, inbox warmup, rotation and cleanup, with campaigns running through Instantly and Lemlist on top. Early on, the domains burned out.
The cause was in the setup. Campaigns were configured manually in Instantly with every tracking marker switched on: open tracking, click tracking, reply tracking.
What fixed it
I turned tracking off. Deliverability improved, and I moved the focus to warming the accounts properly. The live campaign never paused while this happened.
You lose the open-rate number. You keep the inbox. That's the right trade every time.
The rules I run now
The setup grew to 15 domains and 33 inboxes, sending roughly 1,000 emails a day at peak with a reply rate around 5.2%. It ran on a few simple rules:
- Cold sends capped at 30 per inbox per day.
- Warmup starting at 20 to 30 a day, raised in steps of 1 to 2.
- Placement monitoring and gradual volume ramps, so problems show up before a campaign depends on them.
When to stop doing it yourself
I started by warming every inbox myself. Eventually I moved to buying pre-warmed inboxes from Zapmail, because managing it in-house had stopped being worth the time. Cold-email infrastructure companies sell against exactly that decision point. Having been on both sides of it, my answer is to learn it by hand first, so you know what you're paying for when you stop.