# What is a good cold email subject line?
> Specific and boring beats clever. Name the trigger or the topic. Mystery and "Quick question" train filters and humans to skip you.
- HTML: https://omentir.com/help/what-is-a-good-cold-email-subject-line
- Markdown: https://omentir.com/help/what-is-a-good-cold-email-subject-line.md

The subject should still be true after they open. Clickbait that turns into a product dump is how you get deleted and complained on.

Working patterns: a hire, a tool they use, a one-line observation, their company plus a problem. "Acme + AE ramp" is enough. "URGENT partnership" is not.

Length: a few words. Mobile truncates. Front-load the meaning. Lowercase is optional. Personalization tokens that break ("Hi {{FirstName}}") are worse than no token.

Test two subjects on the same list definition. Opens are noisy in 2026 because of privacy protection. Use replies as the real score.

If you cannot write a subject without the product name, you are advertising. Ads can work. They are not what people mean by cold email that gets a conversation.

## Related

- [How do I write a cold email first line?](https://omentir.com/help/how-to-write-a-cold-email-first-line.md)
- [What is a good cold email open rate?](https://omentir.com/help/what-is-a-good-cold-email-open-rate.md)
- [How long should a cold email be?](https://omentir.com/help/how-long-should-a-cold-email-be.md)

## Frequently asked questions

**Should I use Re: or Fwd:?**

No. Fake reply threads are a known trick and a trust hit when they notice.

**Do emojis help?**

They can look like retail promo mail. Skip them for B2B first touch.

**Is the company name in the subject enough personalization?**

It is hygiene. Pair it with a real first line. See [cold email first line](/help/how-to-write-a-cold-email-first-line).

**Can I use the same subject as the InMail?**

You can if it is still true. Do not send both the same morning with the same subject. That is a surround.

[Create an Omentir account](https://omentir.com/signup). Pro is $49/month.
