Email Subject Line Tester

The newsletter took six hours. The photography alone took two. And the subject line — the only part of those six hours most recipients will ever see — got written in eleven seconds at 4:58 p.m. and reads "Quick update + some exciting news!!" It will truncate on every iPhone at the word 'exciting', it triggers two spam heuristics, and the double exclamation point does the exact opposite of what exclamation points are for.

Email remains the channel with the highest ownership and the thinnest margins for carelessness. Your recipients see a sender name, a subject line, a timestamp, and — if you're lucky — a preheader. Everything else, all six hours of it, is behind the click that those few words have to earn. Writing that earns clicks is a craft; checking that the craft survived the last minute of deadline panic is a mechanical task, and mechanical tasks belong to tools.

The Email Subject Line Tester grades any line you paste: length against mobile and desktop windows, spam-trigger vocabulary, casing and punctuation density, emoji cost, personalization presence, and a live truncation preview at common inbox widths. Then it suggests rewrites. It pairs naturally with the Blog Title Generator for the headline muscles, the Meta Description Generator for the search-side cousin of this problem, and the CTR Calculator for measuring what happened after the send.

Below: how to use it, what the score is made of, three lines graded and rewritten, the truncation guillotine, the spam stack, and the preheader pairing that most people forget to write at all.

Take the deadline-panic line above and run it through, because watching a bad line get graded teaches faster than reading rules.

1. 1

   Paste the line

   "Quick update + some exciting news!!" — 35 characters, which sounds safe until the details arrive.

2. 2

   Read the length panel

   35 characters fits the mobile window, but the tester shows where the meaning lands: everything after character 25 is escalating vagueness. 'Update' is doing no work; 'exciting' is a word that has never once made anything exciting.

3. 3

   Check the trigger report

   Exclamation-point density flags — two in a row reads as shouting to filters and humans alike. The plus sign in mid-line is harmless but reads like a template.

4. 4

   See the truncation preview

   The pane shows the line as iPhone Mail renders it, then as desktop Gmail renders it. On the phone the line survives; on the reader's patience it does not.

5. 5

   Read the rewrite suggestions

   Specificity fixes: "Inside: the 3 recipes readers saved most" — 40 characters, a number, a payoff, zero shouting. Same email, honest promise, working verbs.

6. 6

   Score the rewrite

   Paste it back in: the score jumps, the truncation preview still shows the full promise, and the trigger list is empty.

The loop — paste, read, rewrite, re-score — takes about ninety seconds per line, which is roughly 0.4 percent of the time the email itself took. The asymmetry is the whole pitch.

The score is a heuristic — a transparent checklist with weights, not a machine-learning oracle. Transparency is the point: every deduction is something you can see, understand, and argue with.

• Length and window fit. Character count against the classic 40-to-60-character guideline, with preview panes at mobile and desktop widths so truncation is visible rather than theoretical.
• Spam-trigger vocabulary. A weighted dictionary of the classic offenders — free, guarantee, act now, urgency markers, 'click here' — with penalties that stack, because filters score combinations hardest.
• Casing and punctuation. ALL-CAPS words, exclamation-point runs, and suspicious punctuation density, the surface features that make a line look like the emails nobody wants.
• Emoji cost and count. Each emoji counted at its multi-character width with a one-emoji recommendation flag — distinctiveness asset or truncation liability, your call.
• Personalization detection. Merge-field syntax recognized and reported, with the honest caveat that a first name bolted onto a generic blast adds less than senders hope.
• Front-loading analysis. Where the payload sits. The first 35 to 45 characters carry the mobile inbox; a payoff that arrives at character 60 is a payoff most readers never meet.

What the score is not: a prediction of open rates. Open rates depend on sender reputation, list quality, timing, audience fatigue, and a dozen factors no subject-line tool can see. The tester's job is narrower and more useful — catching the mechanical mistakes that quietly bleed opens, before the send makes them permanent. Think of it as the spell-checker for the part of the email that does the selling.

Three realistic lines at three quality levels, with the numbers shown. Scores are illustrative outputs of the heuristic, and the rewrites follow the tester's own suggestions.

The rewrite of the catastrophe writes itself once the deductions are visible: "The 10-minute plan Maya actually finished" — 41 characters, one honest claim, the personalization earns its place, and the promise is one the email has to keep. Note what didn't change: the email still has to deliver. A tester that improves the wrapping without improving the contents just raises the cost of disappointment.

Most email opens on phones, and phone inboxes are narrow. The exact cut varies by device, app, font, and settings — iPhone Mail shows roughly 35 to 41 characters of a subject, Android clients vary more, desktop webmail shows 60 to 70. There is no single safe number, which is why the tester shows previews rather than pronouncing a verdict.

1. Put the payoff in the first 35 characters. Whatever happens after that is a bonus round for desktop readers only.
2. Front-load meaning, not decoration. Emoji, brand prefixes, and bracketed tags — '[Newsletter]' — eat the window before the sentence arrives.
3. Drop the brand prefix on phones if your sender name already says who you are. The name field is free real estate; the subject window is not.
4. Test at both widths. A line that reads perfectly at 70 and becomes gibberish at 35 is a mobile line with a desktop costume on.
5. When in doubt, cut adjectives. 'Exciting new features' is five words of air; 'Offline mode is live' is four words of news.

Character counts also lie the moment emoji enter the line. One emoji renders as a single glyph on screen but measures as two or more characters in the systems that enforce limits, so a subject that looks like 38 characters in your draft can arrive at the inbox measuring 45 — and the truncation point lands three words earlier than the preview suggested. The tester counts what the wire counts, which is the only number that matters when the window is 35 characters wide.

No single word sends an email to spam. Deliverability is a portfolio of sender authentication, engagement history, list hygiene, and content signals — and subject lines are among the most weighted content signals, because they're the part spammers optimize hardest. The tester's model reflects how filters actually behave: individual trigger words are minor deductions, but they compound.

The compounding is the trap. 'Free' alone in a normal sentence from an authenticated domain rarely matters. 'FREE' in caps, followed by '!!!', next to 'guaranteed' and 'act now' — that's not a word, that's a fingerprint, the statistical portrait of a million junk emails. The tester stacks deductions the same way, so a line that leans on one trigger gets a warning while a line that piles four gets the score it deserves. The FTC's CAN-SPAM guidance adds the legal floor: subject lines can't be deceptive about the email's content, which puts fake 'Re:' prefixes and phantom order confirmations out of bounds entirely, deliverability aside.

For the full deliverability picture beyond the subject line — authentication, spam-rate thresholds, list practices — Google's published sender guidelines are the practical reference. The short version: fix authentication first, prune dead addresses second, and then, with the foundation solid, the subject-line polish this tool provides actually shows up in results.

Heuristic scoring catches mistakes. Measuring actual humans requires the second kind of test, and the mechanics are worth knowing even if your platform automates them.

1. 1

   Draft two lines that differ on one axis

   Length, specificity, or personalization — one variable, not three. Two lines that differ everywhere teach you nothing about why the winner won.

2. 2

   Send to a small slice

   Most platforms default to 10 or 20 percent of the list for the A and B arms. Small enough to limit damage, large enough for the comparison to mean something.

3. 3

   Wait for a real interval

   A few hours minimum for B2B lists, less for consumer sends at open-heavy times of day. Declare winners on settled data, not the first ten minutes of enthusiasm.

4. 4

   Roll the winner to the rest

   The platform sends the better performer automatically. Note the margin — a 2-point open-rate difference is noise on small lists; a 6-point difference is a lesson.

5. 5

   Log what you learned

   One line in a doc: audience, both subjects, result. Six months of entries becomes the only subject-line guide that's actually about your list.

The heuristic-to-split pipeline is the honest division of labor. The tester answers 'is this line mechanically sound?' in two seconds; the split test answers 'does this line work on these people?' in two hours. Senders who skip the first step burn their test budget on lines that truncated at character 38; senders who skip the second optimize forever against rules of thumb that their audience quietly stopped obeying.

Newsletter operators with real audiences — the ones whose business survives on opens — test habitually, because they've seen the same line succeed or fail on thirty characters of difference. E-commerce marketers run it against promo subjects, the most trigger-dense genre in email and the one where a stacked 'free shipping ends tonight!!!' costs the most. Nonprofit and community senders, whose emails carry missions rather than discounts, use it mostly for the truncation previews — mission language is wordy by nature and dies quietly at character 41.

The other constituency is everyone writing an important one-to-one email. Job applications, cold outreach, client follow-ups: a subject line that truncates into vagueness on the phone of the person deciding whether to reply. The tester's heuristics were built for marketing volume, but 'Will this line survive the inbox window and read like a human wrote it?' turns out to be a universal question.

The tester grades the subject, but the subject never travels alone. Inbox apps display preview text — the preheader — beside or below the subject line, pulling either from a dedicated preview field or from the first text in your email body. Together they form a two-line advertisement, and the division of labor is clean.

The unforgivable waste is the default: when no preview text is set, clients grab the first body text, which is so often 'View this email in your browser' that the phrase has become visual noise. Setting a deliberate preheader is free — most email platforms expose the field — and it doubles the persuasion surface for zero additional creative work. Write subject and preheader as one thought: subject makes the promise, preheader proves it's keepable, body keeps it.

The tester runs entirely in your browser — the line you paste is tokenized, scored, and previewed by local JavaScript, and nothing transmits anywhere. No account, no email address surrendered as admission, no library of your campaign subjects accumulating on a server you'll never audit.

This matters more than it sounds for a subject-line tool specifically. Unreleased campaign lines are trade secrets — the launch email's subject, the pricing-change announcement's framing. Marketing teams paste pre-send material into tools without thinking, and the habit only becomes visible when a tool's privacy policy gets read on the worst possible day. Local execution removes the category of accident entirely: there is no server-side copy to leak, subpoena, or resell, because there is no server-side anything.

The same property makes it a safe default for teams with approval chains. An agency drafting subjects for a client's product launch can pass the tester's URL around in Slack without a procurement conversation, and a subject line can be reviewed by five people in five tabs without any of them creating an account that marketing operations will eventually have to inventory at renewal time. The boring tools that skip the login wall tend to become the ones that quietly spread through an organization, and a two-second grammar check for subject lines is exactly that kind of boring.

The adjacent pages round out the writing stack: the Blog Title Generator for twenty headline angles when the subject needs siblings, the Meta Description Generator for the 160-character search-facing cousin of this problem, and the CTR Calculator for reading the results afterward. Headlines, subject lines, descriptions — same craft, three windows, one habit of checking the mechanics before hitting send.