Email Subject Lines That Get Opened (and Why Open Rate Lies)
The subject line gets far more attention than it deserves and the sender name gets far less. And since 2021, the metric everyone optimises subject lines against has been substantially broken — which changes how you should test them.
Open rate is not the metric you think it is
Start here, because it changes everything downstream. Apple Mail Privacy Protection, introduced in 2021, pre-loads tracking pixels for Apple Mail users regardless of whether they open the message. Those users are recorded as having opened it.
Apple Mail accounts for a large share of email opens, so a meaningful portion of your reported open rate is now automated pre-fetching rather than human behaviour. Open rates went up across the industry and became substantially less meaningful at the same time.
This does not make subject lines unimportant — they still determine whether real people open. It makes open rate a poor thing to optimise against. Judge subject-line tests on click-through rate and conversions instead, which are still measured honestly.
The sender name does more work than the subject
Before anyone reads your subject line, they see who it is from, and that decides most of it. A recognisable sender gets opened with a mediocre subject line. An unrecognised one gets ignored regardless of how clever the copy is.
Practical implications: keep the sender name consistent so recognition compounds, use a person plus company where it fits your brand, and never send from a no-reply address if you can avoid it. No-reply signals a broadcast, suppresses replies, and replies are a positive deliverability signal.
Deliverability underlies all of it. A brilliant subject line in the spam folder achieves nothing, and inbox placement is driven by authentication, list hygiene and engagement history rather than copy.
Length, and what actually gets truncated
Aim for roughly 30 to 50 characters. Mobile clients show around 35 to 40 before truncating, and mobile is where most email is opened first.
The important technique is front-loading. Put the meaningful words at the start so the subject still works when cut. "Your March invoice is ready" survives truncation; "A quick note about something we wanted to share regarding your account" does not.
Test yours against real client widths with our Email Subject Line Tester, which shows how it renders and flags the patterns that trigger spam filters.
Preview text is the second subject line
Preview text — the snippet after the subject in the inbox list — is the most wasted asset in email marketing. Left unset, clients pull the first text in the email, which is frequently "View this email in your browser" or an image alt attribute.
Set it deliberately on every send. It gives you another 35 to 90 characters and it should extend the subject rather than repeat it. Subject: "Your March invoice is ready". Preview: "Due 31 March, payable in one click." Together they answer more than either does alone.
What works, and what only works once
Specificity beats cleverness. "How we cut a client bounce rate by 40 percent" outperforms "The secret nobody tells you about bounce rate", because it says what the reader gets.
Curiosity works when it is honest. A gap between what the reader knows and wants to know is powerful. A gap the email does not close is a trick, and it converts once before training people to ignore you.
Personalisation beyond the first name. First-name tokens have been used so heavily they now read as automation. Behaviour-based references — what they browsed, what they bought, where they are — still perform.
Avoid: all caps, multiple exclamation marks, "FREE" in capitals, and excessive emoji. These correlate with spam filtering and, more importantly, with being ignored by humans.
Testing it properly
Test one variable at a time or you learn nothing about why a version won. Subject length, or tone, or specificity — not all three at once.
Split your list randomly rather than by segment, or you are measuring the segment rather than the subject line. Make sure the sample is large enough to be meaningful: on small lists, a difference of a few percentage points is noise. Our A/B Test Calculator will tell you the sample size you actually need.
And measure clicks and conversions, not opens. Given the pre-fetching problem, a subject line that wins on open rate and loses on clicks probably did not win at all.
Key takeaways
- Apple Mail Privacy Protection pre-loads tracking pixels, so open rate is substantially inflated and unreliable. Optimise against clicks instead.
- The sender name decides more opens than the subject line. Keep it consistent and avoid no-reply.
- 30 to 50 characters, front-loaded so it survives mobile truncation.
- Set preview text on every send — it is a second subject line most senders waste.
- Specificity beats cleverness; curiosity gaps you do not close train people to ignore you.
Frequently asked questions
What is the ideal email subject line length?
Roughly 30 to 50 characters, because most mobile clients truncate somewhere around 35 to 40 and the majority of email is first seen on a phone. More important than the exact number is front-loading: put the words that carry meaning at the start so the line still works when it is cut off. A longer subject that makes sense in its first six words will outperform a shorter one whose point arrives at the end.
Why did my open rate suddenly increase?
Most likely Apple Mail Privacy Protection rather than anything you did. Since 2021 Apple pre-loads tracking pixels for Apple Mail users whether or not they actually open the message, and those pre-loads are recorded as opens. Because Apple Mail represents a large share of email clients, this inflated reported open rates across the entire industry. If your open rate jumped without a corresponding rise in clicks, that is the explanation, and it means the metric is no longer reliable for optimisation.
Should I use emoji in subject lines?
Sparingly, and only where they fit your brand. A single well-chosen emoji can help a subject line stand out in a crowded inbox, and some audiences respond well to it. Several in one line reads as noise and correlates with spam filtering. The real problem is inconsistent rendering: an emoji that looks right in one client can appear as a hollow box in another. Test before sending, and if you are unsure whether your audience will find it appropriate, they probably will not.
Does personalisation in subject lines still work?
Simple first-name insertion has lost most of its effect, because it has been used so heavily that readers now recognise it as automation rather than attention. What still works is personalisation based on behaviour: referencing a product someone viewed, a category they buy from, their city, or the stage they reached in a process. That is harder to implement because it requires your data to be connected properly, which is precisely why it still stands out.
What is preview text and why does it matter?
Preview text is the snippet shown after the subject line in most inbox lists, typically 35 to 90 characters depending on the client. It matters because if you do not set it, the client pulls whatever text comes first in your email, which is very often "View this email in your browser" or an image alt attribute. That wastes a prominent piece of inbox real estate. Set it deliberately on every send and use it to extend the subject rather than repeat it.
How do I keep emails out of spam?
Deliverability is mostly technical and behavioural rather than about wording. Authenticate your domain with SPF, DKIM and DMARC, which major providers now effectively require for bulk senders. Keep your list clean by removing hard bounces and long-term non-openers, since sending to dead addresses damages your reputation. Make unsubscribing easy, because people who cannot unsubscribe mark you as spam instead, which is far more harmful. Subject-line wording matters least of these, though shouting in capitals still does not help.
How large does my list need to be to A/B test subject lines?
Large enough that the difference you observe is not noise, which usually means at least several thousand recipients per variant for the small percentage differences subject lines typically produce. On a list of a few hundred, a two-point difference in open rate is statistically meaningless and acting on it will mislead you. Calculate the sample size you need before running the test rather than after, and if your list is too small, test bigger differences in approach rather than small wording changes.
Should I still optimise for open rate at all?
Use it as a rough directional signal, not a decision metric. Because a substantial share of reported opens are now automated pre-fetches, comparing open rates between two subject lines can easily point you at the wrong winner. Click-through rate and downstream conversions are still measured honestly and are what you actually care about. A subject line that produces fewer opens but more clicks has attracted the right people rather than everyone, which is usually the better outcome.