Our domain is 11 days old. Here's exactly how we're sending from it.

Run norbelys.com through our own WHOIS & Domain Age tool and you’ll see it immediately: this domain is eleven days old.

That’s usually the kind of fact a company hides. We build deliverability software — surely we should look ancient and established? But the whole premise of this product is that honest numbers beat flattering ones, and an eleven-day-old domain is the most honest test case we’ll ever have. Every receiver on the internet currently treats us exactly the way it treats any new sender: zero history, zero trust, prove it.

So here’s the diary. Everything below is what we actually did, in the order we did it, with the public receipts — because “new domain, day by day” is a path every founder doing outreach walks, usually blind.

Day 1: records before anything

Before the first email — before the website — the authentication records went up, because nothing else matters until they exist:

• SPF, one record, includes only what actually sends for us, lookup budget well under the limit of 10.
• DKIM at 2048 bits — 1024 is on borrowed time, and starting fresh means there’s no excuse to start on borrowed time.
• DMARC at p=none with rua= reporting on from the first hour. Not because p=none protects anything — it doesn’t — but because the reports are the only way to know when we’re ready to climb to quarantine and reject.

You can verify all of it right now: drop norbelys.com into the Domain Health Checker. We don’t get to sell honest tooling and fail our own checks.

Days 2–11: the ramp, not the blast

The temptation with a new product is obvious: announce loudly, email everyone. The math says no. A new domain gets 2–4 weeks of patience minimum, and we’re following the same schedule our Warmup Calculator prints: single-digit daily sends in week one — real emails, to real people who know us — ramping roughly 20–30% at a step, watching bounces against the 2% line, keeping each mailbox inside the human silhouette.

Current status, honestly: small volumes, healthy responses, DMARC reports showing exactly the sources we expect and nobody else. Which is what “nothing to report” looks like, and nothing-to-report is the entire goal of a warmup.

The part that compounds while we wait: giving the checks away

A warming domain can’t do volume, but a website can do work. So days 2 through 11 went into shipping 26 free tools — the checkers, generators and calculators we’d want any sender to run before trusting a platform, no sign-up, running in your browser: blocklist sweeps, header forensics, SPF and DKIM checkers, a DMARC report viewer and the rest — plus one of these posts every day the domain has existed. (You’re reading day eleven.)

There’s no growth hack hiding in that paragraph. It’s the oldest organic play there is: be genuinely useful in public, early, before you’re owed any attention.

What skipping the ramp actually costs

It’s worth spelling out why we didn’t just point our full list at norbelys.com on day one, because “be patient” sounds like caution for its own sake until you’ve watched what happens without it. A brand-new domain has no sending history for any mailbox provider to check — no volume pattern, no engagement signal, no complaint rate to trust. Providers fill that gap with the most conservative assumption available: treat unfamiliar volume as risk until proven otherwise. Send at a normal marketing cadence from day one and the provider doesn’t see “a new company introducing itself” — it sees a domain that appeared from nowhere and immediately started pushing volume, which is also exactly what a freshly spun-up spam operation looks like from the outside. There’s no way for a receiving server to tell the two apart from the traffic pattern alone.

The failure mode that follows isn’t usually a hard block. It’s quieter and harder to notice: a rising share of mail routed straight to the spam folder, with delivery logs still showing “accepted” the whole time. A sender who isn’t watching engagement rates — opens, clicks, replies, not just bounces — can run an entire campaign at effectively zero visibility and never see an error message telling them why. By the time bounce rates climb high enough to notice, the domain’s reputation is already damaged, and undoing that takes longer than building it correctly would have in the first place. The ramp isn’t bureaucracy. It’s the only way to let the data catch up to reality gradually enough that a provider’s spam filter updates its opinion of the domain at roughly the same pace real trust would form anyway.

The mistakes the calendar made easy to avoid

A few decisions in the first eleven days were less about following a formula and more about removing temptation before it had a chance to show up:

• No purchased or scraped list for the first sends. Everything that went out in week one went to people who already know who we are — former colleagues, early testers, people who’d asked to be kept in the loop. Bounce and complaint rates on that kind of list are close to zero by construction, which is exactly what a brand-new domain needs its first real signal to be.
• One sending identity, not several. It’s tempting to spin up a handful of mailboxes on day one “to be ready” once volume picks up. We didn’t — a domain with five freshly created mailboxes all sending small volumes on day one looks stranger to a receiving server than one mailbox building a track record on its own, and every additional identity is one more thing that has to earn trust from zero.
• No content changes to chase deliverability. The instinct when something looks slightly off — a slower-than-expected open rate, a single soft bounce — is to immediately rewrite the email. We resisted that. A domain’s reputation is a function of authentication and sending behavior first; rewriting subject lines in week one mostly adds noise to a signal that’s still too thin to read confidently.

None of these are exotic tactics. They’re closer to just… not doing the things that make a new domain look suspicious, which turns out to be most of what a healthy warmup actually is.

Questions we expect to get asked

What if we needed to send urgently in week one — was patience actually an option? It usually is, more than it feels like. The volumes involved in a genuine first-week send — outreach to people who already expect to hear from you — are small enough that “patient” and “urgent” rarely conflict. The cases where they do are almost always a symptom of treating a new domain like an established one, not an actual scheduling constraint.

How will we know it’s time to move from p=none to p=quarantine? The DMARC aggregate reports are the answer, not a calendar date. Once several consecutive reports show only the sources we expect authenticating cleanly, with no unexplained failures, tightening the policy stops being a guess and starts being a confirmation of what the data already shows.

Does a warmup ramp look different for a brand-new domain versus an established one adding a new mailbox? Yes — a new mailbox on a domain with years of good history inherits some of that domain-level trust, so its ramp can move faster. A domain with zero history, like ours, has nothing to inherit and has to build both domain- and mailbox-level trust from the same starting point, which is why our schedule stays conservative through the full multi-week window rather than accelerating early.

Why publish this at all

Because the advice ecosystem around cold email is allergic to receipts. Everyone says “warm up properly, authenticate, be patient” — from five-year-old domains. Fine: here’s a domain you can carbon-date yourself, following its own advice in public. If it works, the diary is the proof. If something breaks, the diary gets a more interesting entry.

We’ll update as the climb continues — p=quarantine when the reports earn it, real outreach volume when the warmup math says so, actual numbers when there are actual numbers.

Eleven days in, the honest summary: records green, ramp on schedule, nothing to report. Ask us again at thirty.