Outreach
LinkedIn Outreach Email Outreach WhatsApp Automation Dialer Unified Inbox CRM / Pipeline
Data
Signals Lead Finder Email Finder Phone Finder Company Follower
Deliverability
Deliverability Center Mailboundry Email Infrastructure (Google, Microsoft & Azure) Email Warmup Inbox Placement Test
AI & Automation
AI Personalization AI Reply Agent ICP Score
Built For
Founders Agencies Sales Teams B2B SaaS
Use Cases
LinkedIn Outreach Cold Email Outreach Multichannel Outreach Signal-Based Outreach Outbound Sales Lead Generation Account-Based Outreach Appointment Setting Recruiting Outreach Link Building & PR Outreach
Resources
Free Tools Help Center Mobile App API & Webhooks Roadmap Blog Cold Email Guides Deliverability Guides LinkedIn Guides Lead Generation Guides Affiliate Pricing Log in Book a demo Start free trial
BlogDeliverabilityEngage · Deliverability

Deliverability for Non-Technical Founders

You do not need to learn forty acronyms. You need to understand four things and watch three numbers.

RA
← All articles

Deliverability content is written by people who find DNS interesting, for people who find DNS interesting. If you are a founder sending your first few hundred cold emails, almost all of it is more detail than you need and none of it tells you what to ignore.

Here is the short version: you need to understand four things, watch three numbers, and know which problems are worth your afternoon.

The four things

1. Mailbox providers decide, not you

When you send an email, Gmail or Microsoft decides whether it reaches the inbox, the spam folder, or nothing at all. There is no setting you control and no list you can join.

They decide based on two things: do you look like who you say you are (authentication), and do people want your mail (reputation).

That is the whole model. Everything else is detail underneath it.

2. Authentication is three DNS records, set once

Three records prove your mail is really from you. You add them once per domain, at your domain registrar, and then ignore them.

SPF lists which servers may send as your domain. One line. The only trap: it allows a maximum of ten DNS lookups, and stacking several services quietly breaks it.

DKIM signs each message cryptographically so it cannot be altered in transit. Your mail provider generates it; you paste it in.

DMARC tells providers what to do when the first two fail, and asks for reports. Start at p=none, which only monitors. Tighten later.

If you buy mailboxes from an infrastructure provider, these are usually configured for you. Verify anyway, an unauthenticated domain is the most common cause of a cold email programme that never works, and it is a five-minute check. Any free DNS checker will tell you.

3. Reputation is earned slowly and lost quickly

A brand-new domain has no reputation. Not bad, none. Providers treat unknown senders cautiously, which is why your first emails from a new domain land in spam even when everything is configured correctly.

Reputation builds through normal-looking sending over weeks. It falls through complaints, bounces and spam-trap hits, and it falls much faster than it rises. One bad list can undo two months.

The practical implications:

  • Warm up for three to four weeks before real sending. Small volumes, gradually increasing. Most platforms automate it.
  • Never send to a list you have not verified. This is the single highest-risk action available to you.
  • Never move a damaged domain’s problem to a new domain without fixing the cause. You will burn the new one in a fortnight.

4. Volume per mailbox has a ceiling

One mailbox cannot send 500 cold emails a day. The realistic ceiling is roughly 30–50 cold emails per mailbox per day, and new mailboxes should start far below that.

So volume comes from more mailboxes, not more sends per mailbox. If you want 300 emails a day, you need something like six to ten mailboxes, spread across more than one domain. Three mailboxes per domain is sensible; five is the point where you are concentrating risk.

That arithmetic is the whole reason cold email infrastructure exists as a product category.

The three numbers

Check these weekly. They take two minutes and they catch almost everything.

Number Healthy Act at What it means
Bounce rate Under 2% 3% Your list is bad
Spam complaint rate Under 0.1% 0.3% Your targeting or copy is wrong
Reply rate Varies Any sudden fall Often the first sign of a placement problem

Bounce rate is a list-quality problem, almost never a technical one. The fix is verification before sending, not a change to your setup.

Complaint rate has a hard threshold: Gmail enforces at 0.3%, and 0.1% is the number to run at. Three complaints in a thousand emails is a problem. Complaints mean you are emailing people with no plausible reason to hear from you, which is a targeting problem wearing a deliverability costume.

Reply rate falling with no other change is your early warning. Opens are unreliable now, so a quiet reply rate is often how placement decline presents.

When something looks wrong, confirm with a placement test, send to seeded mailboxes across providers and see where it lands, before changing anything. Changing your setup on a hunch is how a list problem becomes a configuration problem too.

What to do yourself and what to pay for

Honest split.

Do yourself (an afternoon):

  • Buy domains. Close variants of your brand, not keyword-stuffed throwaways.
  • Verify authentication with a free checker.
  • Write the sequence. Nobody else knows why your customers buy.
  • Build the first list by hand. Two hundred accounts you chose beats two thousand you filtered.
  • Watch the three numbers weekly.

Pay for:

  • Mailboxes and infrastructure. Setting up and maintaining mailboxes across domains is tedious and consequential. An infrastructure provider charges a few dollars per mailbox per month and removes an entire category of problem.
  • Verification. Cheap, and it prevents the mistake that costs most.
  • The sending platform. Warmup, sequencing, throttling and reply handling are not worth building.

Do not pay for, yet:

  • A deliverability consultant, until you have a measured problem. They are genuinely useful when something is broken and expensive when nothing is.
  • Dedicated IPs, until your volume justifies them. Below roughly 1,500–2,000 sends a month, a shared IP managed by a competent provider gives better results, because a dedicated IP with low volume cannot build a signal. Shared versus dedicated has the threshold.
  • Separate tools for every function. Four subscriptions to learn is four things to get wrong.

The five mistakes that cause most failures

1. Sending from your main company domain. If cold email damages your primary domain’s reputation, your invoices and support email suffer. Use separate domains for cold outreach. Always.

2. Skipping warmup. The most common and the most complete failure. Three weeks feels slow when you want pipeline this month. Skipping it costs more weeks than it saves.

3. Buying a list. Purchased lists contain spam traps, addresses that exist only to catch senders who did not verify. Hitting one is a reputation event, not a bounce.

4. Too many mailboxes on one domain. Ten mailboxes on one domain concentrates all risk in one place. Spread them.

5. Treating a targeting problem as a technical one. High complaints and low replies with clean authentication is not a deliverability problem. It is a message problem. No DNS record fixes a message nobody wanted.

That last one is worth sitting with. Most “deliverability problems” founders bring to me are reply-rate problems, and the sequence is the thing to change.

The ten-minute weekly check

  1. Bounce rate across all campaigns, under 2%?
  2. Complaint rate, under 0.1%?
  3. Reply rate, same as last week?
  4. Any mailbox sending unusually more or less than the others?
  5. Any domain where authentication has changed?

If all five are fine, you are done. If one is not, fix that one thing and check again next week. Changing three things at once means learning nothing.

Engage · Mailboundry

Infrastructure without the admin

Mailboxes on your own domains with authentication configured and warmup built in, shared IP from $0.80 a mailbox, or a dedicated IP for your workspace when volume justifies it.

See Mailboundry

The short version

Providers decide placement based on whether you look authentic and whether people want your mail. Authentication is three DNS records you set once. Reputation builds over weeks and falls in days. Volume comes from more mailboxes, not more sends per mailbox.

Watch three numbers weekly: bounce under 2%, complaints under 0.1%, and a stable reply rate.

Do the domains, the copy and the first list yourself. Pay for mailbox infrastructure, verification and the sending platform. Skip consultants and dedicated IPs until you have a measured reason.

And remember the last mistake: if authentication is clean and replies are poor, the problem is the message.

Next: the cold email audit checklist, or outbound for founders in the first 90 days.

Frequently asked questions

What does a non-technical founder need to know about deliverability?

Four things: mailbox providers decide placement, authentication is three DNS records set once, reputation builds over weeks and falls in days, and volume comes from more mailboxes rather than more sends per mailbox.

How many cold emails can one mailbox send per day?

Roughly 30–50 once warmed, and far fewer during the first weeks. Higher volume needs more mailboxes spread across several domains, with around three mailboxes per domain and five as a practical ceiling.

Should I send cold email from my company domain?

No. Use separate domains for cold outreach so that any reputation damage cannot affect your invoices, support mail and other business email.

Do I need a deliverability consultant?

Not until you have a measured problem, a bounce rate above 3%, complaints above 0.1%, or a confirmed placement failure. Before that, weekly checks on three numbers cover almost everything.

Ready to run outbound on autopilot?

Start free trial