Email Subject Lines: Write Clear Lines Without Deceiving

The two ways subject lines fail#

Subject lines fail in one of two ways. The first is being forgettable: "Our latest products" gets deleted unread. The second is being deceptive: "Re: your order" on a message that is not a reply, or "70% off" when the email contains no discount. The second failure is worse, because a materially misleading subject line is a per-email violation under CAN-SPAM, and because it trains your audience to ignore you.

Write the body first#

A subject line is a summary of an email that already exists. If you write the subject before the body, you are guessing. Write the body, then ask: "in one short line, what is the single most important thing in this email?" That line is your subject. The discipline of summarising forces you to have an actual point.

Respect the character budget#

Many recipients read email on mobile, and subject-line display length varies by client and device. Put the key noun early so the important part survives shorter previews. Keep the key noun — the product, the offer, the question — in the first 25 characters, and use 50 characters as a practical target, not an absolute limit.

Nine subject lines that follow the rules#

These examples lead with a useful noun and should match the email they introduce. Check each length and mobile preview before sending:

1. "The 1.7L kettle is back in stock"
2. "Your saved basket expires tonight"
3. "5 winter layers, one price drop"
4. "Order 1042 has shipped"
5. "How your bottle survived the drop test"
6. "3 colours back: sage, slate, sand"
7. "A 2-minute fix for a leaking lid"
8. "Your warranty: what year 2 covers"
9. "We answered your sizing question"

Notice what is absent: no "Re:" on a non-reply, no fake urgency, no percentage the body cannot honour. The persuasion comes from specificity, which is also what keeps each line truthful.

The one-line summary test#

Before sending, imagine a recipient forwarding your email to a friend with the subject line pasted above it. Does the combination make sense, or does the subject promise something the body does not deliver? If the latter, one of them is wrong. This test catches both deception and vagueness in one pass.

The sender name is half the subject line#

Shoppers decide to open (or delete) based on the pair — sender name and subject. A recognisable sender ("Aurora Home", not "no-reply@aurora") buys your subject line the fraction of a second it needs. A from-address that changes every send, or a name in ALL CAPS, undoes the best-written subject. Pick one sender name, keep it, and never send marketing from a no-reply address — replies are signal, and CAN-SPAM's honest-identification requirement makes deceptive headers a per-email violation anyway. The FTC's CAN-SPAM compliance guide covers the full set: no misleading headers, a working unsubscribe honoured within 10 business days, and a valid physical postal address.

What genuinely works#

• Specificity: "The 1.7L kettle is back in stock" beats "We have news".
• Curiosity with a payoff: "Why your water bottle leaks (and the 2-minute fix)".
• Numbers that are real: "5 settings, 3 colours, 1 decision".
• The recipient's own context: "Your basket is saved" to someone with an abandoned cart.

What to measure — and what not to trust#

Open rate is a blunt instrument now that privacy tools pre-load tracking pixels; treat it as a trend line, not a number. The truer signals are click rate (did the subject's promise survive contact with the body?) and unsubscribe and complaint rate (did the subject over-promise?). A subject line that lifts opens but lifts complaints even faster is losing you the list.

The pre-send checklist#

• Put the key noun early; aim for roughly 50 characters when practical, but treat that as a readability target rather than an absolute platform limit.
• Every claim in the subject is delivered by the body — discount, stock, deadline.
• No "Re:" or "Fwd:" unless it is genuinely one.
• Sender name matches the one your list already knows.
• Unsubscribe works, and the physical address is in the footer.

Write the body, summarise it honestly, keep the sender consistent — the humble version of email marketing that keeps working while cleverer versions get marked as spam.

The claim rules follow you into the inbox#

Most subject-line advice talks about spam filters as a technical problem — trigger words, caps, punctuation. That is the shallow layer. Underneath it sits the same claim law that governs listings and ads, because a subject line is an advertisement for the email.

• Fabricated urgency — "last chance", "ends tonight", "only 3 left" — is only legal while it is true. A deadline that resets every week is a deceptive-practice pattern that regulators have repeatedly acted on.
• Guarantee language — "risk-free", "100% satisfaction guaranteed" — must match a policy you actually honour, with its material limits visible.
• Outcome promises — "lose 5kg", "cure your joint pain" — move the email into health-claim territory, where the substantiation bar is the same as for a product page.

This is why serious email tooling treats claim scanning as part of the send, not an afterthought. We run email copy through the same ruleset as listings — the rules do not care that the channel is an inbox; the policies the rules derive from certainly do not.

Three rewrites, annotated#

Before: "🔥 LAST CHANCE — 100% FREE trial, ends TONIGHT!!!" After: "Your trial ends Friday — here is what is in the paid tier" Why: real deadline, no absolutes and no fake urgency. Measure the result with your own audience rather than assuming a performance lift.

Before: "Cure back pain with this one exercise" After: "A 10-minute back-care routine" Why: removes the unsupported cure and invented testimonial. Verify the routine and review medical wording before sending.

Before: "Re: your order" (on an email that is not a reply) After: "About the order you were browsing" Why: fake "Re:" framing can mislead recipients and may create deliverability or consumer-protection risk. Use reply framing only when the message is genuinely a reply.

None of these rewrites are less persuasive. They are less punishable — by filters, by law, and by the reader's memory of having been tricked.