The B2B Cold Email Deliverability Checklist: What to Check and When

Quick summary: Cold email deliverability is whether your message reaches the primary inbox, and it depends on authentication, domain and mailbox reputation, sending volume, list quality and copy. Most teams know the theory. What breaks campaigns is that nobody checks the right thing at the right moment. This checklist is organized by when each check belongs: before a domain sends, before a list loads, before a sequence goes live, during launch week, every week, and the day replies drop.

Why a Checklist, and Not Another Explainer

Most B2B teams spend hours on copy, subject lines and sequence logic. Far fewer check whether any of those emails actually land in the primary inbox. That gap between effort and outcome is where outbound pipeline quietly disappears.

A broken deliverability setup never announces itself with an error. Emails go to spam, to the promotions tab, or get silently discarded by a corporate gateway. Reply rates fall, the team assumes the message is wrong, and the sequence gets rewritten. The infrastructure underneath never gets touched.

If you want the background on why this happens, we have already covered it in depth: why your cold emails land in spam walks through how providers decide, and inbox placement vs. open rates explains why the metric most teams watch can look healthy while the inbox is lost. This page does not repeat that. It is the operational version: a list of yes-or-no checks, grouped by the moment in a campaign's life when each one actually matters.

How to Use This Checklist

Treat every line as a pass or fail, not a "mostly". A partial SPF record, a list that was "probably" verified, or a mailbox that "should be" warm is a fail. Deliverability problems compound, and one weak link is enough to drag down a whole campaign.

• Give each section an owner. Someone has to be accountable for DNS, someone for list quality, someone for weekly monitoring. On small teams that may be one person, which is fine, as long as it is written down.

• Run the checks at the moment they belong to. A pre-send check run after launch is a post-mortem.

• Record the result. A dated note of what was checked and what was found is what lets you diagnose a drop three weeks later instead of guessing.

Stage 1: Before a Domain Sends Its First Email

This is the foundation. Everything that happens later inherits the decisions made here, and most of them are cheap to get right and expensive to fix after a domain has a damaged reputation.

1. Cold outreach runs on dedicated sending domains, not your primary business domain. If a campaign goes wrong, the damage stays on a domain built for outreach, and your invoices, client threads and support mail keep landing.

2. Each sending domain has exactly one SPF record, and it includes every service that sends on that domain's behalf.

3. DKIM is signing outgoing mail, and you have confirmed it on a real sent message, not only in the DNS panel.

4. DMARC exists with a reporting address. Starting at p=none and tightening later is a reasonable path, as long as someone reads the reports.

5. SPF and DKIM align with the visible From domain. Alignment is the part that breaks silently. Our SPF, DKIM and DMARC setup guide covers how to verify each record and the order to set them up in without breaking anything.

6. MX records exist and the mailboxes can receive. A sending domain that cannot receive replies looks suspicious to providers and loses you the replies you worked for.

7. Click tracking, if you use it at all, runs on a custom tracking domain. A shared tracking domain lends you the reputation of every other sender on it. Many teams turn tracking off entirely for cold outreach and lose nothing that matters.

8. The domain and its mailboxes have real sending history before cold volume starts. Either they have been warmed, or you are using pre-warmed infrastructure. Brand-new mailboxes sending cold at volume on day one is the most common self-inflicted wound in outbound.

9. Volume is spread across several domains and mailboxes. No single sending identity should carry the whole campaign, so one bad week on one domain does not take everything down with it.

Stage 2: Before a List Goes Into a Sequence

List quality is the input your sender reputation is built on. Hard bounces and spam complaints are among the fastest ways to damage a domain, and almost all of them can be prevented before the first send.

1. Every address has been through an email verification tool within a short window before sending, not months ago when the list was bought or built. B2B data decays as people change roles.

2. Catch-all domains are handled deliberately. Decide whether you send to them, in what proportion, and from which mailboxes, instead of letting them in by default.

3. Role addresses such as info@, sales@ and support@ are removed unless you have a specific reason to write to a shared inbox.

4. The list is checked against your blocklist and do-not-contact records, at both email and domain level. Someone who asked their whole company to be removed should not be reached again through a colleague.

5. Contacts are deduplicated across active campaigns, so the same person is not hit by two sequences in the same week.

6. The list matches your ICP. Irrelevant recipients do not reply, and a pattern of silence and deletions is itself a negative signal. Relevance is a deliverability factor, not only a conversion one.

Stage 3: Before a Sequence Goes Live

Spam filters read the message, not only the headers. These checks take minutes and catch problems that no DNS setup can compensate for.

1. The first email is plain text or close to it. No images, no heavy formatting, no layout that looks like a newsletter.

2. The first email carries one link at most, and ideally none. A first touch only needs a reply, not a click.

3. The subject line describes what is actually in the email. Subject lines that promise something the body does not deliver train recipients to mark you as spam.

4. The copy avoids bulk-marketing language such as "act now", "guaranteed" or "no obligation". It should read like one professional writing to another.

5. Merge fields fail safely. Preview the sequence with an empty first name and an empty company field, and make sure nothing reads "Hi ,".

6. There is variation across sends. Thousands of identical bodies from the same infrastructure form a fingerprint. Light, grammatical variation in phrasing helps every message look like what it is: an individual email.

7. There is a clear way to opt out, whether a link or a plain "reply no and I will not follow up". It keeps you compliant and it is cheaper than a spam complaint.

8. A seed or inbox placement test has been run from the actual sending mailboxes with the actual copy. Tools such as GlockApps or Mail-Tester show where the message lands across major providers and flag authentication or content problems before a real prospect sees anything.

Stage 4: Launch Week

The first days of a campaign are when problems are cheapest to catch, because they have not compounded yet. Watch closely and do not scale until the numbers have settled.

1. Daily volume per mailbox starts conservatively and only rises once bounces and replies look healthy. The right ceiling depends on domain age, warm-up history and engagement, so set it from your own data rather than a number you read online.

2. Bounces are reviewed per mailbox and per domain, daily. A campaign-wide average hides the one mailbox that is in trouble.

3. Every reply is read and categorized, including out-of-office messages, wrong-person redirects and removal requests. Replies are your clearest live signal that mail is landing.

4. Removal requests reach the blocklist the same day. Do not wait for an automation to catch up. A second email to someone who asked to be removed is how complaints happen.

5. Any mailbox showing a sudden spike in bounces is paused while you find out why, instead of being allowed to keep sending.

Stage 5: Every Week, for as Long as the Campaign Runs

Deliverability is not a setup task. Reputation moves every week based on how recipients and providers react to what you send. A short recurring review catches drift before it turns into a collapse.

1. Reply rate is tracked per domain and per mailbox, not only per campaign. A domain that used to get replies and suddenly does not is the earliest warning you will get.

2. Hard bounce trend is tracked, and any rise is traced back to a specific list source or segment.

3. Blacklist status is checked for every sending domain and its IPs, using a lookup tool such as MXToolbox or Spamhaus. Some listings cause silent throttling instead of visible bounces, so a campaign can look like it is running while it is not.

4. DMARC reports are read, or at least scanned for unexpected senders and failures.

5. Warm-up and health status of every mailbox is reviewed. Disconnected, blocked or flagged mailboxes get fixed or replaced, not left in rotation.

6. Your do-not-contact categories and your blocklist still match. Gaps here are how people who asked to be removed get emailed again.

7. Open rate is not used as a health signal. Privacy features and security scanners trigger opens automatically, so it tells you little about real inbox placement. Replies and bounces are the honest numbers.

Stage 6: The Day Replies Drop

When results fall with no change in the copy, the instinct is to rewrite the sequence. Resist it until you have worked through this order. It goes from the most likely and cheapest-to-check cause to the least.

1. Run a fresh inbox placement test from the affected mailboxes. If mail is landing in spam or promotions, the problem is reputation or infrastructure, not the message.

2. Check bounces for the period of the drop. A new list source with poor data is the most common hidden cause.

3. Check blacklists for the affected domains.

4. Check whether anything changed in DNS. A new tool added to a domain, a migrated provider or an edited SPF record can quietly break authentication or alignment.

5. Check volume. Did someone raise daily limits or add a large new campaign onto the same mailboxes?

6. Only then look at the copy and the targeting. If placement, bounces, blacklists, DNS and volume are all clean, the message or the audience is the real issue, and now a rewrite is the right move.

When a domain is badly damaged, recovery means lower volume, cleaner lists and patience, and it can take a long time. Sometimes the practical answer is to retire the domain and move the campaign onto healthy infrastructure, which is another reason to never let a single domain carry everything.

Who Owns This Checklist: Platform-Led or Done-for-You

Every item above has to be done by someone. The question is who.

In a Platform-Led setup, your own team runs campaigns on Lidgen's platform and infrastructure. The foundation stages are handled for you: campaigns launch on pre-warmed domains and mailboxes, so there is no month-long warm-up and first leads can arrive within 24 hours. Your team owns the list, copy and reply checks, with the infrastructure underneath already built to pass Stage 1.

In a Done-for-You setup, a partner runs strategy, infrastructure and execution, and your team approves the messaging direction. The whole checklist, including the weekly reviews and the reply-drop triage, becomes the partner's daily routine.

Both models run on the same foundation: pre-warmed infrastructure and human-in-the-loop review of every message. The choice is about bandwidth and how hands-on your team wants to be, not about how safe the result is. That foundation is how Lidgen campaigns reach near-100% primary inbox placement, and it is why so many of the checks above are about people reviewing signals that automated systems tend to misread.

Frequently Asked Questions

What is a safe daily sending volume per mailbox for B2B cold email?

There is no universal number. It depends on the age and history of the domain, how well the mailbox was warmed, and how recipients are responding. Start conservatively, watch bounces and replies per mailbox, and raise volume only when both look healthy. Spreading volume across several domains and mailboxes is safer than pushing any single one.

How do I know if my cold emails are landing in spam or the primary inbox?

Run an inbox placement or seed test from your real sending mailboxes with your real copy. It shows which folder each provider puts the message in. In live campaigns, a drop in replies with no change in copy is the most practical warning sign.

What bounce rate should worry me?

Any sustained rise. Track hard bounces per mailbox and per list source, and treat an upward trend as a list or infrastructure problem to trace immediately rather than an acceptable cost of volume.

Does email copy affect deliverability, or is it purely technical?

Both. Filters read content alongside authentication. Heavy formatting, several links, promotional phrasing and misleading subject lines can hurt placement even when every DNS record is correct.

Can I send cold outreach from my primary business domain?

It is not recommended. If a campaign damages the domain's reputation, all of your business email suffers, including client communication. Keep cold outreach on dedicated sending domains.

What is the difference between a soft bounce and a hard bounce?

A hard bounce is a permanent failure, such as an address that does not exist, and the contact should be removed immediately. A soft bounce is temporary, such as a full inbox or a server issue. An address that keeps soft-bouncing should be treated like a hard bounce and removed.

How often should I run this checklist?

Stages 1 to 3 run every time you add a domain, a list or a sequence. Stage 4 runs during every launch. Stage 5 runs weekly for as long as campaigns are live, and Stage 6 runs the moment results drop without an obvious reason.