Landing in the inbox instead of spam

"Delivered" and "in the inbox" are different things. A message can be accepted and still be filed in spam, and nothing in your own reporting will tell you — the only reliable check is looking at a real inbox on the provider your recipients actually use.

In rough order of impact:

1. Authenticate your domain

Non-negotiable, and the largest single factor. An unauthenticated sender is treated with suspicion regardless of content. See Verify a sending domain.

2. Keep complaints near zero

The tightest constraint you operate under — 0.08% is about one in twelve hundred. See Spam complaints.

3. Send to people who want it

Engagement is what providers measure now. Opens, clicks and replies from real people are the evidence that you are wanted; a list where nobody interacts drags down delivery for the recipients who do.

Mailing your engaged contacts and leaving the dormant ones alone usually improves placement more than any change to the email itself.

4. Keep bounces low

See Bounces, and why they matter so much.

5. Be consistent

A steady monthly send beats a burst of five emails after a year of silence. Providers read sudden volume changes from a quiet sender as a compromised account, because that is often what it is.

6. Make the email look like mail

• Include a plain-text version.
• Do not send one large image with no text — a classic spam signature, and invisible to the many recipients whose client blocks images.
• Make the unsubscribe link easy to find. Hiding it produces complaints, which cost far more than the subscriber would have.

What will not fix it

Rewording the subject to avoid "spam trigger words" is largely folklore. Modern filtering is driven by authentication, reputation and engagement, not by whether you wrote "free". A well-authenticated sender with an engaged list can write "free" all day.

Next: open Senders and confirm your domain reads Verified — that is the first thing to fix, and the one with the biggest effect.