What the Email Subject Line Generator does
You describe the email — what it's about, who it's going to, what you want them to do. The widget returns 5 to 10 subject line variations, each written in a different pattern (curiosity, value, urgency, question, direct). You pick the one that fits.
The premise: most people write one subject line, slap it on the email, and send. That's the line you came up with in three seconds at the end of writing the body. A subject line is the only thing standing between your email and the archive button — it deserves more than three seconds. Generating five options and picking the best takes about thirty.
This isn't a guessing tool. It's a "give me options so I can choose deliberately" tool. The model doesn't know which line will perform best for your specific audience — nobody does, until you A/B test. But it'll write five lines that each have a reason to exist, and one of them is almost always better than what you'd have shipped on autopilot.
How to use it
- In the input box, describe the email's purpose. Be specific: "sales follow-up after a product demo" is much better than "follow-up email."
- Add context the subject line should reflect: the audience (prospects, existing customers, internal team), the urgency (now-or-never vs evergreen), the key number or hook if you have one.
- Hit Generate. The widget returns 5–10 variations.
- Read them out loud. Cut the ones that feel like spam. Pick the one that sounds like something you'd open.
- Copy and paste into your email client.
One thing to remember: the subject line preview gets cut off on mobile around 40 characters. About 60% of business email is opened on a phone now. If the strongest part of your subject line is at the end, it won't be seen. Keep the punch in the first 40 characters and treat anything after that as bonus.
Worked example: sales follow-up after a demo
You ran a product demo with a prospect yesterday. The conversation went well — they had three good questions and asked about pricing at the end. Now you're writing the follow-up. The intent input reads:
"Sales follow-up after yesterday's demo with the head of ops at a mid-size logistics company. Want to send the pricing one-pager and propose a 15-minute next call. Friendly but professional."
Five subject lines come back, each in a different pattern:
- Curiosity: "One thing I forgot to mention yesterday"
- Value: "Pricing + the 15-min next step you asked about"
- Urgency: "Demo recap (and your Q on pricing) before the week gets away"
- Question: "Quick follow-up — did Tuesday's demo answer the routing question?"
- Direct: "Demo follow-up: pricing one-pager attached"
Five different bets on what gets a busy ops leader to click. The curiosity line is risky but high-ceiling; the direct line is safe but low-ceiling. For a warm prospect who already met you, the value or direct lines are usually the right call — they're specific, they reference what was discussed, and they signal that the body is worth the click. The curiosity line works better for cold outreach where you have nothing to reference back to.
You'd never have written five lines manually. You'd have written "Demo follow-up" and moved on. The widget's job is to make picking deliberately as cheap as picking lazily.
The five subject-line patterns, and when each one works
The widget cycles through five core patterns because nearly every high-performing subject line is a variant of one of them. Knowing the patterns helps you pick the right variation faster.
| Pattern | How it works | Reported open-rate lift vs generic | When it fails |
|---|---|---|---|
| Curiosity | Withhold the key piece of info; promise it inside | +22% (cold outreach studies) | If the body doesn't deliver, you train the recipient to ignore you |
| Value | State the concrete thing inside the email | +14% (B2B SaaS data) | Becomes generic if the "value" isn't specific |
| Urgency | Time-based reason to open now | +18% (e-commerce) | Fake urgency burns trust fast; only works when real |
| Question | Direct question, expects a mental answer | +10% (sales outreach) | Yes/no questions perform worse than open ones |
| Direct | Just states what the email is | +5% over no subject; baseline | Boring, but rarely misfires |
The numbers above are averages across published industry studies (Mailchimp, HubSpot, Litmus, Yesware). Your specific audience and list will vary. The point isn't that curiosity always wins — it's that any of these patterns will outperform "Re: Touching base" or "Quick question."
The 40-character rule
Most mobile email clients show 30–50 characters of subject line before truncating. Gmail on iOS displays around 36 characters in portrait. Apple Mail shows about 40. Outlook varies by device but is usually in the 40–60 range.
The implication: the first 40 characters of your subject line are the entire signal for over half your readers. Anything after that is a footnote at best, invisible at worst. "Following up on yesterday's conversation about Q3 pricing" reads as "Following up on yesterday's convers…" on mobile, which is a generic enough line to get archived.
The widget's outputs are calibrated to land under 50 characters where possible. If a variation runs long, the strongest hook is placed early. When you're picking between options, count the characters (or just count the words — six words is a useful rough ceiling) and prefer the shorter one if all else is equal.
What makes a subject line bad
The widget's outputs are usable. The widget's failure mode — and the failure mode of email subject lines in general — is reaching for desperate moves that read as spam. Watch for these in any variation, AI-generated or hand-written:
- ALL CAPS. Reads as shouting. Gets filtered by spam classifiers. Never opens at higher rates than the sentence-case equivalent.
- Excessive emoji. One can work as a visual anchor in a crowded inbox. Three or more triggers spam filters and signals low-effort sender. The widget will sometimes emit a single emoji — that's fine. Delete any beyond one.
- "Re:" or "Fwd:" when there's no thread. A short-term open-rate gain, a long-term trust loss. Recipients flag it and your sender reputation drops. Don't do it.
- "Free" and "!!!" stacked together. Classic spam triggers. Modern filters don't auto-reject these, but they noticeably increase the spam-folder probability.
- Vague check-ins. "Touching base," "circling back," "quick question," "checking in." These have no information content. Open rates on these phrases have dropped year over year for a decade.
If a generated variation contains any of the above, regenerate or pick a different one. The model occasionally over-leans on these because they appear constantly in training data — not because they work.
How to A/B test subject lines (if you can)
For one-to-one email — sales outreach, recruiter messages, networking — you can't really A/B test. You send one subject line per recipient. The discipline is just to pick deliberately rather than reflexively.
For email campaigns — newsletters, marketing blasts, transactional sequences — A/B testing is the only honest way to know which pattern works for your audience. Most email platforms (Mailchimp, ConvertKit, Klaviyo, HubSpot, Customer.io) have built-in split testing. The protocol that works:
- Generate 3–5 variations from the widget
- Pick the two that feel most different from each other — usually one direct, one curiosity
- Split your list 50/50, send, measure open rate after 48 hours
- Use the winner as the sent line for the rest of the list (most platforms automate this)
Two cautions. First, you need at least ~1,000 recipients per variant for the open-rate difference to be statistically meaningful — smaller lists produce noise. Second, open rate has become less reliable as a metric since Apple's Mail Privacy Protection started inflating opens in 2021. Reply rate or click-through rate is a sturdier signal if your sample is large enough to measure them.
Cold outreach vs warm follow-up: different rules
The same email purpose can call for very different subject lines depending on whether the recipient knows who you are.
| Context | Best pattern | Why |
|---|---|---|
| Cold outreach to a stranger | Curiosity or question | You have no relationship equity. The subject is the entire pitch for opening. |
| Warm follow-up after a meeting | Value or direct | The recipient knows you. The subject just needs to be specific and useful. |
| Newsletter to existing list | Curiosity or direct | Open rate on newsletters lives or dies by the subject — the body is almost irrelevant to the open decision. |
| Internal email to coworker | Direct | The recipient will open it regardless. Just label the content clearly so they can triage. |
| Transactional (receipt, password reset) | Direct | Functional. Curiosity here is anti-pattern; the user is looking for the email and needs to find it. |
The widget doesn't ask which bucket you're in, so the context lives in your input. "Cold email to a CTO at a fintech startup" produces different outputs than "follow-up to the CTO I met at a conference last week." More context in the input, better-targeted variations in the output.
Related tools
- Email Generator — the natural pair. Generate the subject line, then generate the body it sits on top of.
- Caption Generator — same pattern (short attention-grabbing text), applied to social posts instead of email.
- Hashtag Generator — for the social-distribution side of any marketing email you're also posting elsewhere.
- Product Description Generator — if your email is selling something, the subject line and the product description are working together.
Frequently asked questions
How many variations should I generate?
Five is usually plenty. Past about seven, the variations start repeating patterns and the marginal new option doesn't add much. If none of the first five feel right, regenerate with a more specific intent rather than asking for more variations on the same vague brief.
Do emojis in subject lines help or hurt?
It depends on your audience and how sparingly you use them. One emoji can act as a visual anchor in a packed inbox and lift open rates by 2–5% in some consumer-list studies. Two or more usually hurt — they read as marketing-blast energy and trigger more aggressive spam filtering. For B2B audiences, default to no emoji; for B2C and creator-audience email, one emoji can earn its place.
What's the ideal subject line length?
Six to ten words, or about 40–50 characters. Long enough to be specific, short enough to fit in mobile previews. Subject lines shorter than four words can feel cryptic; longer than 12 words start to read as a sentence rather than a label.
Should I capitalize every word?
No. Sentence case ("Quick follow-up on yesterday's demo") consistently outperforms title case ("Quick Follow-Up On Yesterday's Demo") in tested data. Sentence case reads as human; title case reads as a marketing department. The widget outputs sentence case by default.
Does the subject line matter if my open rate is already high?
Yes — but the lever changes. If you're at 50% open rate, you're not pulling people in cold; you're maintaining a base. The subject line job shifts from "earn the open" to "set the right expectation for the body." A misleading subject on a warm list erodes the relationship over time even if it boosts opens this week.
Can I save subject lines I like?
The widget doesn't save state between sessions. Copy the lines you want into a doc — many marketing teams keep a swipe file of strong subject lines they've used or seen, sorted by pattern. It's a useful reference when you're stuck.
Does this work for languages other than English?
You can write your input in any language the underlying model supports — Spanish, French, German, Portuguese, Japanese, and others. The variations come back in whatever language you wrote the input in. The character-count rules shift across languages (German runs longer for the same idea, Japanese runs shorter), so check the mobile preview length for your specific language.
How does this compare to spending $59/month on a subject-line AI service?
Those services run on the same underlying language models the rest of the industry uses, wrapped in a subscription paywall. The $59 covers their marketing budget and their dashboard — not better output. Microapp's AI features run at near-cost; members get a monthly credit allotment included, and beyond that it's pay-per-use at roughly what the model API costs. The output quality is competitive with the paid services; the price difference is one of the cleanest examples of paying for branding versus paying for the actual thing.