Why do emails land in spam and how do you write a subject line that does not?
What actually triggers filters, and how to test subject lines instead of guessing.
Most email lands in spam because of who you are sending to and how they react, not because you used the word free. Reputation, authentication and engagement decide placement. Subject line wording is a smaller factor, though some patterns still hurt.
The real causes, in order
- Complaint rate. Google's sender guidelines tell senders to keep the spam rate in Postmaster Tools below 0.10 percent and never reach 0.30 percent. Yahoo publishes the same 0.3 percent ceiling. This is the fastest way to lose the inbox.
- Missing or broken authentication. No SPF, DKIM or DMARC, or a From domain that does not align with them.
- Bad list. Purchased addresses, spam traps, hard bounces you kept mailing.
- No engagement. Mailing people who never open or click teaches the filter that your mail is unwanted.
- Volume spikes. Google warns that suddenly doubling your usual volume can trigger rate limiting and reputation drops.
- Then content. Image-only emails, link shorteners, mismatched link domains, and sloppy HTML.
Subject lines, honestly
No single word sends you to spam by itself. What hurts is the pattern: all capitals, stacked exclamation points, unrealistic money claims, urgency that is not real, and the visual shouting that spam has trained filters to recognize. Avoid the pattern rather than memorizing a banned word list. We keep a working reference in our list of subject line spam triggers.
Write for the person instead. Say the specific thing. Front-load the useful word, because mobile truncates. Make the preview text carry real information instead of repeating the subject. And do not promise something the email does not deliver, because the complaint you earn costs more than the open you bought.
Test with enough people to mean something
A subject line test on 200 contacts tells you nothing. If the difference between two variants is a couple of percentage points and your sample is small, you are reading noise. Test on your largest sends, test one variable at a time, hold the send time constant, and judge on clicks rather than opens, because open tracking is no longer reliable. Keep a running log of what won, since the pattern across twelve tests is worth more than any single result. Method is in our A/B testing guide.
Related: Can LevelUp Digital build custom HubSpot workflows and automations?. What HubSpot services does LevelUp Digital offer?.