the one-root-domain system that runs 60 inboxes off a $10 domain, tested across 30,000 accounts holding 99.7% health.
Before You Buy Another 20 Domains
you're about to buy 20 domains to stand up 60 new inboxes. that's roughly $200 gone, and you'll do it again next quarter when they burn.
there's a version of this where you spend $10.
after setting up cold email infrastructure over 2,000 times in the last 12 months, and testing this exact system across about 30,000 accounts, i can tell you the domain bill is the single most overspent line in outbound. one root domain can carry 60 inboxes. most people are paying for 20.
this is the contrarian one everyone's been told doesn't work. the received wisdom says subdomains are "all bound together" by the receiving providers, so if one goes down they all go down. we tested it at scale. the data says the OPPOSITE.
what you'll walk away with:
- the root-and-branch mechanism that lets one $10 domain carry up to 60 sending inboxes
- the exact build spec we run internally, subdomains, inbox counts, DNS and daily send config
- the test data behind it, including a 5,000-account client that sent 1.8 million emails at 99.7% inbox health
- the cost model that turns a $2,000 domain bill into a $10 one, and a burned domain into a $12 non-event
- the four places you should NOT plant a tree, so you don't take a good idea somewhere it breaks
- a one-page build cheat sheet you can hand your ops person
you can run this on your next batch of inboxes, on the exact structure our biggest accounts use right now.
let's cut your domain bill by 90%.
The Root-and-Branch Myth
the myth you've been sold is simple. every sending domain has to be its own separate purchase, because subdomains are "all bound together" by the receiving providers. flag one, the theory goes, and Google takes down the whole family.
so everyone buys dedicated domains. 20 domains to stand up 60 inboxes. $200, gone the moment they burn.
mate, we tested that assumption across about 30,000 accounts. it doesn't hold.
picture your root domain as a tree. we NEVER send from the root. the root just sits there, clean, holding everything up. all the sending happens on the branches, the subdomains hanging off it. and the part the myth gets backwards is this... when a branch burns, we cut it off, and the tree lives. one subdomain getting flagged does not take down the other nineteen.
send from the root domain itself and one bad batch WOULD burn the whole tree. that's the real failure mode people are scared of. they're just pointing it at the wrong part of the tree.
so the "subdomains are bound together" panic only bites when you send from the root. keep the root dark, and the branches isolate a burn on their own.
that single shift, root stays clean and branches do the sending, is what lets one domain carry 60 inboxes at the same performance as 60 separate ones. the rest of this blueprint is how you build it.
Cost Your Domain Bill in 60 Seconds
before anything else, price what you're overspending on domains right now. it takes a minute.
grab three numbers from your current setup:
- how many sending inboxes are you running? (call it your inbox count)
- how many domains are those inboxes spread across?
- what did each domain cost you? (a .com is usually $10 to $12)
now run the numbers on the back of a napkin.
the subdomain way, one root domain carries up to 60 inboxes. so the domains you ACTUALLY need is your inbox count divided by 60, rounded up. for most people that comes out at one.
worked example. say you run 300 inboxes. the old way at 2 to 3 inboxes per domain, that's 100 to 150 domains, so $1,000 to $1,800 in domain spend. the subdomain way, 300 inboxes needs 5 root domains, about $60. you're overspending by more than $1,700 on domains alone, every cycle they burn.
write yours here:
your inbox count: ______ domains you're paying for now: ______ (x ~$10 = $______) root domains you'd actually need, inboxes / 60: ______ (x ~$10 = $______) what you're overspending: $______
if that overspend number stung, the rest of this shows you how to claw it back without losing a point of performance.
Build the Tree: One Root, Sixty Inboxes
this is the structure that carries 60 inboxes on one domain. build it exactly like this.
start with ONE root domain. a clean .com from Porkbun, Namecheap or GoDaddy, forwarded through Cloudflare. this root is the trunk. it never sends a single cold email. its only job is to hold the branches.
off that root, you spin up subdomains. these are your branches, and they look like send.yourdomain.com or mail.yourdomain.com. you go up to 20 branches on one root, and no higher. that ceiling matters, we do not push past 20 per root, because that's where the isolation stays reliable.
on each branch, you build 2 to 3 inboxes. official Google Workspace or Microsoft 365 accounts, built fresh on the subdomain, never cracked or farmed. 20 branches at 3 inboxes each lands you at 60 sending accounts on a single $10 root.
THE BUILD SPEC (what we run internally):
- 1 root domain, dark, never sends
- up to 20 subdomains (branches) per root
- 2 to 3 official inboxes per branch
- SPF, DKIM, DMARC and MX set correctly on every branch before it sends a single email
- 60 inboxes maximum per root domain
now the daily send config, because this is where people cook their branches. on the reference client running this at scale, each inbox sends 10 cold emails a day plus 20 warm-up emails a day, 30 total per box. the warm-up runs in parallel with the campaigns, not before them, so the providers never see a behaviour jolt.
one honest adjustment i make on my own accounts. 60 boxes per root is the ceiling, not the target. my biggest clients run about 10 branches, so 30 accounts per root. it's a mental barrier number more than a hard rule. i take the upper limit and i run HALF, because the extra headroom keeps the whole tree calmer.
key takeaway: keep the root dark, cap it at 20 branches, and 2 to 3 fresh inboxes per branch gives you 60 accounts on a $10 domain.
The Test Data
before you trust a structure with your pipeline, you want the receipts. we have them.
we finished testing subdomains as a strategy across about 30,000 accounts, and the performance held basically the same as dedicated domains. not close-enough-same. the SAME. that's exactly why this blueprint exists, the data refused to break.
the headline proof is one reference client. 5,000 accounts, all built on the root-and-branch structure, 10 subdomains per root, 30 accounts per domain, official Google. over six weeks they sent 1.8 million cold emails. the numbers came back like this:
- 99.7% average inbox health across all 5,000 accounts
- 0.6% bounce rate
- 0.4% positive reply rate on 1.8 million sends
- 2,000+ opportunities generated
read those against the fear. the myth says stacking 5,000 accounts on shared roots should have collapsed deliverability. instead the health score sat at 99.7. that client has since DOUBLED their infrastructure and now runs around 10,000 accounts on the exact same structure.
an email account doesn't just combust on a timer. it degrades from a combination of things, usually volume and sending behaviour together. the subdomain structure was never one of them. build the branches right and keep the root dark, and the tree holds.
key takeaway: 30,000 accounts tested and a 5,000-account client at 99.7% health over 1.8 million sends... the structure holds at scale.
The $10 Domain Bill
this is where the structure pays you back. two numbers make the case.
the build cost first. on dedicated domains you need about 20 domains to stand up 60 inboxes, and at $10 a domain that's $200 before you've sent an email. the subdomain way, you need ONE domain for the same 60 inboxes. $10. same performance, a twentieth of the domain spend.
now the part that changes how you feel about burning. every cold email setup burns eventually, that's the game, not a failure. the question is what a burn costs you.
on dedicated domains, when a domain burns you're replacing a $200 chunk of infrastructure. on subdomains, when a branch burns you cut it off and replace ONE domain. you're $12 in the hole instead of $2,000 in the hole. across 10 domains it's $12 versus $120.
that's the reframe. a burned branch stops being a disaster and becomes a $12 line item you barely notice. you can run more aggressively, keep more failover ready, and push more volume, because the downside of a burn just went from painful to trivial.
there's an upsell hiding in this too. because the domain saving is so large, you can resell it as volume. for an extra $100 or so in domains, you can double a client's sending capacity. you go from the forecast you quoted to double it, on $10 domains. why would they say no?
key takeaway: one root domain runs 60 inboxes for $10, and a burned branch costs you $12 instead of $2,000.
When NOT to Plant a Tree
this system is not a free lunch with no rules. take it somewhere it doesn't belong and you'll burn a whole tree instead of a branch. four places to be careful.
first, NEVER send from the root domain. i know it's said three times in this guide now. that's on purpose. the root is the one thing keeping the branches independent. send one campaign from it and a single burn takes down all 60 inboxes at once. keep it dark, always.
second, don't stack a cheap TLD on a subdomain. a .info domain performs fine as a root at a quarter of the .com price, but i won't run .info ON a subdomain. that's shaky on shaky. the branch structure is already the aggressive play, so build it on clean .com roots and don't stack extra risk on top.
third, respect the 20-branch ceiling. 60 inboxes per root is the maximum, not the target. on my own accounts i run about 10 branches, 30 accounts per root, half the ceiling. if you max every root at 60 to save a few dollars, you've traded your safety margin for pennies. take the upper limit and run half.
fourth, if you're small, the extra setup might not be worth it yet. running 30 inboxes total? the subdomain saving is real but small, maybe $190, and it's more moving parts than one flat domain. this is a play for scale. it earns its complexity when you're running hundreds or thousands of accounts, not your first thirty.
key takeaway: keep the root dark, don't stack .info on branches, cap at 20, and only reach for this once you're scaling.
The Subdomain Build Cheat Sheet
pin this above your monitor and build straight off it.
THE STRUCTURE:
- 1 root domain (a clean .com), forwarded through Cloudflare, NEVER sends
- up to 20 subdomains (branches) per root, and no more
- 2 to 3 official Google or Microsoft inboxes per branch
- 60 inboxes maximum per root, though 30 (10 branches) is the comfortable setting
THE DAILY CONFIG (per inbox):
- 10 cold emails a day
- 20 warm-up emails a day, run in parallel with campaigns
- 30 total per box
- SPF, DKIM, DMARC and MX green before the first send
THE COST:
- 60 inboxes on dedicated domains: ~20 domains, ~$200
- 60 inboxes on subdomains: 1 root, ~$10
- a burned dedicated domain: ~$200 gone
- a burned branch: replace 1 domain, ~$12
THE RULES:
- never send from the root, that's what keeps the branches independent
- don't stack .info on a subdomain, shaky on shaky
- cap at 20 branches, run half where you can
- it's a scale play, reach for it in the hundreds and thousands, not your first thirty
THE PROOF:
- tested across ~30,000 accounts, performance held the same as dedicated domains
- 5,000-account client: 1.8 million emails, 99.7% inbox health, 0.6% bounce, 0.4% positive reply
build the tree, keep the root dark, and let the branches do the work.
