Building Scalable Business Infrastructure: What Running a Million Cold Emails Taught Me
- May 6
- 2 min read
By Jack Kelly
Most founders treat infrastructure as something you build after the business takes off. That's backwards. The systems you put in place when you're small are the ones that determine whether you can grow without everything falling apart.
I run MyColdLeads with one other person. We're a cold email agency based in Seville, Spain, helping B2B companies across Europe generate qualified sales meetings through personalized outreach. Between us, we've sent over a million emails for clients across industries from education to immigration law to industrial software. The only reason two people can do this is because we built the infrastructure first.
Start with the boring stuff
The systems that matter most aren't exciting. Ours started as checklists - a shared doc with steps we followed before every campaign launch. How we verify email addresses before sending, because a single bad bounce can damage your sender reputation for weeks. How we warm up new accounts for 14 days before using them. How we rotate domains on a fixed schedule to keep everything healthy.
None of this required fancy software. We just wrote down what we were actually doing and stuck to it. The moment we started skipping steps because we were busy, things broke. A client's campaign tanked because we compressed the warmup timeline. That taught us that repeatable processes aren't optional.
Automate the repetitive parts, keep the thinking human
Once we had reliable processes documented, we automated the parts that didn't need a human brain. We use n8n to handle the repetitive steps: validating email addresses, monitoring domain health, and flagging problems before they affect live campaigns.
That freed us to focus on work that actually moves results - understanding what makes a specific prospect respond. We test four different subject lines and four email bodies on every campaign, because the data tells us things our instincts get wrong constantly.
The real infrastructure problem
The operational mistake that costs the most time isn't "we don't have a process." It's "we have five different processes and nobody knows which one to follow." When you're small, consolidation matters more than sophistication. One system that everyone actually uses beats three elegant solutions that nobody trusts.
We learned this with our email validation. We tried multiple tools independently, each with their own workflow. Results were inconsistent. So we built a single validation waterfall - one process that runs every address through multiple checks in sequence and gives a clear pass or fail. Simple, boring, works every time.
Build for the team you want, not the team you have
We built our systems assuming we'd need to hand them to someone new eventually. Every process is documented. Every automation has clear inputs and outputs. If something breaks at 2 AM, the fix shouldn't require tribal knowledge.
That discipline is what lets two people in Seville serve clients across a dozen countries without burning out. The systems do the heavy lifting. We do the thinking.
Connect With Jack
Comments