Deliverability

Your domain didn't get bad at email. It got punished.

Your domain didn't get bad at email, it got judged. The real reasons emails land in spam and a short checklist to get back to the inbox.

A glowing red envelope above a laptop, surrounded by warning triangles

Short answer

Emails go to spam when mailbox providers stop trusting the sending domain. The usual causes are missing or failing authentication (SPF, DKIM, DMARC), too many spam complaints, bounces to addresses that do not exist, sudden jumps in volume and low engagement. Fix authentication first, then stop sending to unverified and unengaged addresses so the reputation can recover.

Nothing changed. Same newsletter template, same sending tool, same Tuesday morning slot. And yet the open rate halved, and a colleague forwarded a screenshot of your campaign sitting in their Gmail spam folder.

It is tempting to blame the subject line, or the new image, or a word that sounds salesy. Usually it is none of those. Your domain did not get bad at email. It got judged, and the verdict applies to everything it sends.

Spam placement is a verdict on the sender

Gmail, Outlook and Yahoo decide where a message goes mostly by asking one question: do we trust whoever sent this? The answer comes from months of history attached to your sending domain and the IP addresses that send for it.

That is why a perfectly ordinary message can land in spam, and why moving one campaign around does not fix it. The filter is not reading the email in isolation. It is reading your record.

The five things mailbox providers actually judge

1. Authentication. Can the provider confirm the mail really comes from your domain? That is what SPF, DKIM and DMARC are for. Large senders must have all three: Gmail and Yahoo require it for bulk senders, and Gmail moved to rejecting non-compliant mail outright rather than only filtering it. Microsoft rejects failing high-volume mail to Outlook addresses with a 550 5.7.515 error.

2. Complaints. Every "Report spam" click counts against you. Google asks bulk senders to keep the rate reported in Postmaster Tools below 0.1% and never to reach 0.3%. That is one complaint in a thousand messages.

3. Bounces. Mail to addresses that do not exist tells providers you do not know your audience. Why bounces burn more than a campaign explains how a few hundred dead addresses spread into everything else.

4. Volume patterns. A domain that sends 500 emails a week and then 50,000 on a Tuesday looks compromised or careless. Sudden spikes invite suspicion even when the list is clean.

5. Engagement. Opens, replies, moving mail out of spam, adding you to contacts: positive signals raise trust. Mail that is ignored, deleted unread or left in spam lowers it.

The part content plays

Content matters less than most people think, but it is not irrelevant. Misleading subject lines, link shorteners, a single image with no text and links to domains with poor reputations all add weight on the wrong side.

On a domain with a strong record, ordinary marketing copy rarely decides anything. On a domain already under suspicion, it can be the last straw. Fix the record first.

Four explanations that waste a week

Before rewriting the newsletter, rule these out. They are the usual suspects, and they are usually innocent.

  • "A word in the subject line triggered a filter." Modern filters weigh sender history far above vocabulary. The word free in a subject line from a trusted domain is fine. The same word from a domain with a 4% bounce rate is not the reason it was filtered.
  • "Our image to text ratio is wrong." A single large image with two lines of text is a weak signal, not a verdict. It matters at the margin, on a sender who is already borderline.
  • "The unsubscribe link makes us look like a marketer." The opposite. A visible, working unsubscribe link is required for bulk senders and prevents the complaints that actually hurt.
  • "We should switch sending platform." Domain reputation travels with the From domain, not with the tool. A new platform changes the IP and nothing else, and on a shared pool it can make things worse.

The common thread: content decides edge cases, reputation decides the rest.

Spam folder, or never delivered at all?

These look the same from a sender's chair and need opposite fixes.

Filtered means the message was accepted and then placed in spam. Your platform reports it as delivered, the open rate collapses, and someone eventually finds it in a spam folder.

Blocked means the receiving server refused it. It appears in your bounce log with a 5.7.x code and wording about policy, reputation or a blocklist. Nobody will ever find it, because it was never stored.

SymptomLikely stateNext step
Delivered, near-zero opens at one providerFilteredCheck authentication and complaint rate
550 5.7.1, 554 5.7.1 in the bounce logBlocked by policy or reputationRead the full text of the rejection, it often names the reason
The rejection mentions a blocklist URLListedCheck the domain reputation and IP reputation tools, then request delisting
Delivery delayed by hours, then arrivesThrottledReduce volume per hour and keep it steady

How to find out which one it is

  • Look at the headers of a message that landed in spam. spf=pass, dkim=pass and dmarc=pass should all be there. A fail is your first fix.
  • Check Google Postmaster Tools for your domain's spam rate and compliance status with Gmail's sender requirements.
  • Check the blocklists. Run your domain through the domain reputation checker and your sending IP through the IP reputation checker. A listing explains sudden rejections.
  • Read your bounce report per campaign and per list source. One source that bounces far more than the others is often the whole story.

Read the headers yourself

Every message carries the receiving provider's own verdict on your authentication. It takes a minute to read and settles most arguments.

In Gmail, open the message, choose "Show original" from the menu. In Outlook on the web, open the message properties. Look for the Authentication-Results line:

Authentication-Results: mx.google.com;
  dkim=pass header.d=example.com;
  spf=pass smtp.mailfrom=bounce.example.com;
  dmarc=pass (p=NONE sp=NONE dis=NONE) header.from=example.com

Three things have to be true:

  1. dkim=pass, and header.d is your own domain, not your sending platform's. A signature from sendgrid.net passes, but it does not build your reputation.
  2. spf=pass, for the domain in the envelope sender.
  3. dmarc=pass, which requires SPF or DKIM to be aligned with the domain your recipients actually see in the From field.

Send one test to a Gmail address and one to an Outlook.com address before any campaign goes out. If any line says fail, softfail, none or temperror, stop here: nothing else you change will be judged fairly until this reads pass.

Getting back to the inbox

Order matters:

  1. Fix authentication. Nothing else sticks while SPF, DKIM or DMARC fail.
  2. Stop sending to addresses you have not verified. Run your lists through verification and remove invalid addresses. How to reduce your email bounce rate covers the three moves.
  3. Send to your most engaged contacts first. People who opened or clicked recently rebuild trust fastest.
  4. Raise volume gradually over a few weeks instead of returning to full blasts on day one.
  5. Make unsubscribing effortless, so annoyed recipients leave instead of complaining.

The habit that prevents most of it

Most spam-folder stories start with a list nobody checked. Keep this short:

  • Verify addresses when they are collected, so disposable and mistyped ones never enter. Disposable email detection shows the signup side.
  • Verify any list older than a few months before sending to it.
  • Suppress hard bounces and complaints automatically, everywhere.
  • Watch spam rate and bounce rate every month, not only when something breaks.

The fastest way to protect a sender reputation is to stop giving providers reasons to doubt it. Bulk verification cleans the list you send to next.

FAQ

Questions people ask about this

Why do my emails go to spam in Gmail but not in Outlook?

Each mailbox provider scores senders on its own data. Gmail may have seen complaints or bounces from your domain that Outlook has not, or its users may engage differently with your mail. Check each provider's signals separately rather than assuming one result applies everywhere.

Can the content of an email alone send it to spam?

It can contribute, especially misleading subject lines, link shorteners or a single large image with no text. But on an established, trusted domain, ordinary marketing copy rarely decides placement. Reputation, authentication and complaints usually matter far more.

How do I know if my domain is on a blocklist?

Look the domain and your sending IP up against the major blocklists. BounceIntel's free domain reputation and IP reputation checkers query them without signup. A listing explains sudden rejections and usually points at a specific sending problem to fix.

How long does it take to get back into the inbox?

Usually a few weeks of careful sending. Fix authentication, remove invalid and long-inactive addresses, send to your most engaged recipients first, and raise volume gradually. Providers restore trust as they see consistently good behaviour.

How do I tell whether my email went to spam or was blocked?

Check the send log. Filtered mail is reported as delivered but never opened, while blocked mail appears as a bounce with a 5.7.x code mentioning policy, reputation or a blocklist. They look the same from the outside and need opposite fixes.

Do spam words in the subject line cause emails to go to spam?

Rarely, on their own. Modern filters weigh sender history far above vocabulary, so the same wording is fine from a trusted domain and irrelevant as an explanation for a domain with high bounces or complaints. Fix authentication and list quality first.

Protect your sender reputation before the next send.

Verify a list, wire the API into your signup form, and keep bad addresses away from your campaigns.