Learn

Outbound infrastructure & deliverability

Make sure the emails you send actually land in the inbox, this month and next.

Roadmap: Outbound engine: infrastructure & copy

Why it matters

The best list and the best copy are worth nothing in a spam folder. Deliverability is invisible when it works and catastrophic when it fails: a burned primary domain can disrupt a whole company's ordinary email, including invoices and customer support. Since Google and Yahoo tightened bulk-sender rules in 2024 (with Microsoft following), authentication, low complaint rates and sane volumes are requirements rather than best practices. GTM engineers own this plumbing at many companies.

What it's made of

  • Secondary sending domains and how they relate to the main brand
  • SPF, DKIM and DMARC: what each proves and how to check them
  • Mailbox warmup and gradual volume ramp
  • Per-inbox and per-domain sending limits
  • Bounce, spam-complaint and placement monitoring
  • Tracking choices: custom tracking domains, and why open rates mislead
  • Unsubscribe handling and the basics of cold-email law by region
  • Multichannel limits: LinkedIn and calling alongside email

Authentication in plain words

Three DNS records tell receiving servers that your mail is really from you:

  • SPF lists which servers are allowed to send mail for your domain.
  • DKIM adds a cryptographic signature to each message so receivers can confirm it wasn't altered and really came from your domain.
  • DMARC tells receivers what to do if SPF or DKIM fail (none, quarantine or reject) and where to send reports. Start with p=none to collect reports, then tighten.
  • Check all three with a DNS lookup tool before any inbox sends a single email, and recheck after any change to your sending tool.

Capacity planning: work backwards

Say the team needs 40 meetings a month. If about 1 in 100 emails sent leads to a meeting in your market, that's around 4,000 new-contact emails a month, plus follow-ups. A three-step sequence roughly triples sends to about 12,000, or about 550 per working day.

At a conservative 30 emails per inbox per day, that's around 18–20 inboxes. With 2–3 inboxes per domain, you need 7–10 secondary domains, plus a few spares warming up so you can rotate. Doing this arithmetic before buying anything prevents both under-capacity and the urge to over-send from too few inboxes.

Warmup and ramp

New domains and inboxes have no reputation. Warm them for two to four weeks with low, gradually rising volume before cold sending, and keep warmup running at a low level afterwards. Ramp real campaigns slowly too: start each new inbox at 10–15 cold emails a day and increase week by week while watching bounces and replies.

What to watch every week

  • Hard bounce rate per campaign: keep it under about 2%. Above that, stop and fix the list.
  • Spam complaint rate: the large mailbox providers want it under 0.3%, and well below 0.1% is a healthier target.
  • Reply rate trend per domain: a sudden drop on one domain while others hold steady usually means placement trouble there.
  • Inbox placement tests across Google and Microsoft mailboxes before and during big sends.
  • Unsubscribes honoured quickly across every tool, not just the one where the person replied.

Mid-level looks like

Sets up secondary domains with SPF, DKIM and DMARC, warms up inboxes, and keeps each one under a conservative daily limit.

Senior looks like

Designs sending capacity from pipeline targets, monitors placement and complaints per domain, rotates or rests domains before they degrade, and can diagnose a deliverability drop from headers and data.

Common pitfalls

  • Sending cold email from the company's main domain.
  • Buying 50 inboxes and sending 100 emails each on day one.
  • Judging campaigns by open rate, which privacy features and security scanners have made unreliable.
  • Ignoring bounces and complaints until reply rates have already collapsed.

Practice drills

  1. Drill 1

    Plan the sending infrastructure for a team that wants 60 meetings a month. Show your conversion assumptions, inbox and domain counts, warmup timeline and the date the first cold email can go out.

    A strong answer: Arithmetic from meetings back to daily sends, conservative per-inbox limits, spare domains, a realistic warmup period, and assumptions labelled.

  2. Drill 2

    Set up SPF, DKIM and DMARC on a test domain, send to a mailbox you control and read the authentication results in the message headers.

    A strong answer: All three passing, an explanation of each header line, and a DMARC policy plan from monitoring to enforcement.

  3. Drill 3

    Reply rates on two of your eight domains fell from 4% to under 1% in a week while the others held steady. Write your diagnosis steps and what you'd change.

    A strong answer: Isolating the affected domains, checking bounces, complaints, authentication and placement, pausing those domains, and fixing the cause rather than adding volume elsewhere.

  1. 1Ultimate Guide to GTM Engineering for 202613 min
  2. 2Replace Your SDR Team With Clay: Step-by-Step Clay Outbound Masterclass77 min

Best resources for this skill

Repos, courses, articles and podcasts, ranked by how much they teach.

Read next

Next skillCopywriting & personalisation at scale