You just raised. You're about to hire SDRs. Here's how teams like yours burn their domain in 90 days — and how to not.
Whether it's your first outbound motion or you're scaling past 10,000 sends a day with AI agents in the loop, the failure mode is the same: sending before the infrastructure exists. I build it so you never find out what recovery costs.
The 90-day domain burn
It's not bad luck and it's not bad copy. It's the same five infrastructure mistakes, in roughly this order. I see this pattern weekly.
Sending cold email from the root domain
The fastest way to take down your own company email. When yourcompany.com gets flagged, it takes investor updates, customer onboarding, and password resets down with it. Cold volume never belongs on the domain your business runs on.
No warmup — volume on day one
New domains and mailboxes have zero sending history. Gmail and Microsoft treat sudden volume from a cold identity as a compromised account or a spammer. Proper warmup takes 14–30 days; most teams give it zero.
Volume spikes that look like account takeover
Going from 50 to 500 sends a day overnight because "the SDRs started Monday" is exactly the pattern mailbox providers are trained to flag. Volume has to ramp on a schedule filters recognize as organic.
Unverified list data
Scraped or stale lists bounce. Sustained bounce rates above 2–3% read as negligence to mailbox providers, and a single spam-trap hit can put a fresh domain on a blocklist before the first reply comes in.
AI-generated content tells, at scale
AI SDR tools write fast — and they write in patterns. Gmail and Microsoft in 2026 are demonstrably better at clustering templated AI text across thousands of inboxes. Personalization architecture, not just prompts, is what keeps agent-written email out of spam.
The architecture I build
Seven layers, each with a job. This is what "outbound infrastructure" actually means when an engineer builds it instead of a checklist.
Secondary sending domains
Cold volume lives on dedicated lookalike domains — never your root. Sized by the math: roughly 2–3 domains per SDR at healthy volume, named so replies still feel like your brand.
Mailbox provisioning
2–3 mailboxes per domain, capped around 20–30 cold sends per mailbox per day. Provisioned on infrastructure chosen for your stack — Google, Microsoft, or dedicated providers.
Authentication, enforced
SPF, DKIM, and DMARC configured and aligned on every sending domain from day one — the 2024+ Gmail/Yahoo bulk-sender requirements treat this as table stakes, not extra credit.
Warmup orchestration
14–30 day engagement-tiered warmup before a single sequence goes out, with ramp schedules that match how mailbox providers expect a real sender to grow.
Sequencer configuration
Instantly, Smartlead, Apollo, lemlist — connected with correct rotation, daily caps, reply handling, and tracking-domain setup. The defaults are where deliverability goes to die.
Monitoring
Google Postmaster Tools, Microsoft SNDS, blocklist watch, and inbox placement testing — so reputation drift gets caught in days, not after reply rates crater.
Incident response runbook
When a domain degrades — and at scale, eventually one will — your team has the documented playbook: what to pause, what to rotate, who to delist with, and how to re-ramp.
Read the full guide: The GTM Email Infrastructure Stack (2026)
How engagements work
Audit first. You'll know exactly what's broken — or missing — before you spend a dollar on ongoing work.
Outbound Infrastructure Audit
I map your domains, authentication, mailbox setup, sequencer configuration, and reputation standing. You get a findings report and a prioritized build plan — whether or not we work together after.
Outbound Infrastructure Buildout
The productized build: secondary sending domains, mailbox provisioning, SPF/DKIM/DMARC, 14–30 day warmup, sequencer configuration, monitoring, and a runbook your team keeps.
Managed Deliverability Retainer
I run it: continuous monitoring, reputation management, incident response, and new domain provisioning as you scale. Bounded scope — you know exactly what is included.
Questions GTM teams ask
How many domains and mailboxes do I need per SDR?
The working math: one SDR sending 150 cold emails/day needs roughly 6 mailboxes (at 20–30 sends per mailbox per day), spread across 2–3 secondary domains. An AI SDR agent doing the volume of several humans needs proportionally more. I size this precisely during the audit based on your volume targets.
Can I send cold email from my main company domain?
No. Cold outbound belongs on secondary domains that are isolated from your root domain. If a secondary domain gets flagged, you rotate it out and your company email keeps working. If your root domain gets flagged, investor updates and password resets go to spam too.
How long before my SDRs can actually start sending?
Plan for 3–5 weeks from kickoff: about a week for domain and mailbox provisioning plus authentication, then 14–30 days of warmup before sequences launch at ramped volume. Teams that skip this are the ones burning domains in the first 90 days.
Does this work with AI SDR tools like 11x, Artisan, or AiSDR?
Yes — AI SDR agents need this infrastructure more than human SDRs, not less. They send at higher volume with machine-generated content, which is exactly the combination spam filters in 2026 are best at catching. The architecture is the same; the volume math, content variation, and monitoring thresholds change.
What does outbound email infrastructure cost?
The buildout is a one-time $5,000–$10,000 depending on team size and volume targets, plus roughly $50–$300/month in domain and mailbox costs you pay directly to providers. Ongoing managed deliverability runs $3,500–$4,500/month. The audit ($1,500–$2,500) comes first and prices the rest precisely.
We already have outbound running. Is it too late to fix the foundation?
No — this is the most common starting point for scaling teams. The audit maps what you have against what your volume requires, and the buildout migrates sending onto sound infrastructure without pausing your pipeline. If reputation damage already exists, that's diagnosed and priced as part of the same audit.
Not ready to talk? Audit yourself first.
The free Deliverability Audit Checklist covers the same checks I run in paid audits — authentication, reputation, and sending patterns.
Ready? Let's talk about your infrastructure.
Book a free 30-minute strategy call. You'll walk away with a clear diagnosis and a plan of action — whether you hire me or not.