How to Write Email Subject Lines People Actually Open

A subject line has one job: get the right person to open the email for the right reason. It isn't there to be clever or to win a copywriting award. The best subject lines read like a note from someone the reader trusts, telling them exactly what's inside.

Below: how subject lines and preview text work together, which "spam words" are real risks, how to test honestly without built-in A/B testing, 23 before-and-after rewrites, and a pre-send checklist.

What the subject line and preview text are each responsible for

In most inboxes a reader sees three things before deciding: the sender name, the subject line, and a snippet of preview text. Treat them as a single unit with a division of labour.

• Sender name answers "who is this from, and do I care?" Use a name people recognise, and keep it consistent from send to send.
• Subject line answers "what is this about?" It should name the topic or the offer in plain words.
• Preview text answers "why should I open it now?" It adds the detail, the benefit, or the deadline the subject line had no room for.

If you leave the preview text empty, many email clients fill the space with the first text they find in the email body. Writing the preview text deliberately is one of the cheapest improvements you can make.

A common mistake is to repeat the subject line in the preview text. You're wasting the second line. If the subject says "Our spring catalogue is here," the preview should add something new, such as "12 new pieces, including the linen jacket you asked us to bring back."

How much of your subject line people actually see

No single character limit is correct, because the visible length depends on the email client, the device, the screen width, the font, and even whether the reader has turned on a reading pane. A subject that shows in full on a desktop monitor can be cut off halfway on a phone.

So instead of chasing a magic number, write for the worst case:

1. Put the meaning first. The first three or four words should carry the message on their own. "Workshop seats close Friday" survives truncation. "Hey there, just a quick reminder that workshop seats close Friday" doesn't.
2. Test on your own devices. Send a test to a Gmail account, an Outlook account, and an Apple Mail account, and check each on a phone. That takes five minutes and shows you what your readers really see.
3. Treat the preview text as a continuation. If the subject gets cut off, the preview text should still finish the thought.

Clarity versus curiosity

Curiosity subject lines ("You won't believe this", "Quick question") can work well, but only when the sender has already earned the reader's trust. For a newer list, or for a reader who skims forty emails before breakfast, clarity almost always wins. A clear subject line tells people what they get, and the people who open it are the ones who want it.

Curiosity also has a cost you won't see in an open count. If the email doesn't deliver on what the subject line hinted at, readers learn to ignore you, or they mark you as spam. Gmail's email sender guidelines say a message's subject, headers, and display name should accurately represent the sender and the content. They also warn against subject lines that start with "Re:" or "Fwd:" when the email isn't actually a reply or a forward.

A useful rule: curiosity is fine, deception isn't. "The mistake we made with our pricing page" is curious and honest, because the email really is about that mistake. "Re: your order" on a promotional email is deceptive.

Formulas that stay on the right side of that line

• Outcome + timeframe: "Plan a month of content in one afternoon"
• Specific number: "3 changes that cut our support tickets"
• Plain announcement: "New: gift cards are live"
• Honest question: "Still using the old export?"
• Deadline with a real date: "Early pricing ends Thursday at midnight"

Personalization that comes from real data

Putting a first name in the subject line is the version of personalization everyone knows, and it's also the weakest. Readers are used to it, and when your data is messy it goes wrong in public: "Hi FIRSTNAME" or "Hi jsmith88."

Stronger personalization uses something you actually know about the reader:

• What they bought or signed up for: "Care tips for your new cast-iron pan"
• Where they are: "Our Leeds pop-up opens Saturday"
• Their plan or role: "For agency owners: client reporting in 10 minutes"
• Their last action: "The guide you downloaded, part two"

This only works if the data is clean and stored where your email tool can use it. In Cresca you can import contacts from a CSV with column mapping, then store attributes such as city, plan, or product interest as tags and custom fields. Build a segment from those fields and write a subject line just for that group. That's usually more effective than a merge tag, and it's much safer.

Always set a fallback for any merge field. If a contact has no first name, the subject should still read naturally.

Spam filters: what matters and what's a myth

A lot of subject-line advice still comes from an era when filters scanned for banned words. Modern mailbox providers rely much more on who you are and how recipients react to you. Look at what Gmail and Yahoo's sender best practices actually ask for: authentication (SPF, DKIM, and DMARC for bulk senders), a working unsubscribe, and a low spam-complaint rate. Neither publishes a list of forbidden subject-line words.

If your emails land in spam, look at authentication and complaints before you rewrite your subject lines. Our guide on why emails go to spam walks through the order to check things.

Testing subject lines without built-in A/B testing

Many platforms offer A/B testing that splits a list and picks a winner by open rate. Cresca doesn't include that feature, and that's worth being upfront about. Open rate is also a shakier judge than it used to be: privacy features in some mail apps load images automatically, which inflates opens. We explain the details in why open rates stopped being reliable.

You can still learn which subject lines work. You just test more slowly and judge by outcomes you trust.

A simple manual testing method

1. Pick one variable. For example, a question versus a statement, or naming the price versus not naming it. Change nothing else.
2. Split into comparable segments. Divide a list into two groups that are as similar as possible, such as alternating sign-up months or a random split you make in your spreadsheet before importing. Tag each group.
3. Send version A to one group and version B to the other, at the same time on the same day.
4. Measure what matters. Count replies in your inbox. In your store or site analytics, compare sales or sign-ups from links tagged with a different utm_content value for each version (a UTM tag your analytics reads from the link). In your campaign report, check unsubscribes and complaints for each version.
5. Repeat before you conclude. One send proves very little. Run the same kind of comparison across three or four campaigns and look for a pattern.

Keep a simple log with the date, segment, subject line, preview text, and outcomes. After a few months, that log will teach you more about your audience than any generic list of "best subject lines."

Before-and-after examples

These rewrites are illustrative. They show the reasoning, not results from real campaigns.

If you write with AI, generate several options, then edit them against the checklist below. AI is good at variety and bad at knowing what's true about your business. Cresca's AI template generator drafts the email and its copy from a prompt, but you still decide which line is honest and specific enough to send.

The pre-send subject line checklist

• Do the first few words carry the message even if the rest is cut off?
• Does the preview text add new information instead of repeating the subject?
• Is the sender name recognisable and the same as last time?
• Does the email deliver exactly what the subject line promises?
• No fake "Re:" or "Fwd:", and no emoji pretending to be buttons?
• Does every merge field have a fallback, and did you test it with a contact who has missing data?
• Is any deadline real, with a date or time?
• Did you check how it renders on at least one phone and one desktop client?
• Is the sending domain authenticated with SPF and DKIM?
• If you're testing, did you change only one thing, and do you know which outcome you'll measure?

Subject lines improve through habit, not tricks. Write plainly, keep your promises, and log what your readers respond to.

Continue learning

Related Cresca resources

• Email marketing with Cresca
• Explore Cresca features

External references

• Gmail email sender guidelines
• RFC 8058: one-click unsubscribe