agentleFS
manypicommanypicom/sales-skills/skills/subject-line-optimization/SKILL.md

subject-line-optimization

Skill1 starsChanged 2 months agoMITView on GitHub

Write and test cold email subject lines that get opened without misleading the recipient. Use whenever the user mentions subject lines, open rates, email headlines, preview text, preheaders, or wants subject lines written or reviewed. Also use when open rates are low, or when someone wants to A/B test the opener of a cold email.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Plain: No punctuation marks, no emoji, no brackets, no "RE:" on a first email.

    line 30
  2. Distinctive:

    Use them sparingly and always with a fallback.

    line 48
  3. Distinctive:

    If you're going to test, do it properly or don't do it.

    line 90
  4. Distinctive:

    Don't optimise subject lines: when reply rate is the problem.

    line 136

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Subject Lines

A subject line has one job: get the email opened by someone who might genuinely care. Getting it opened by everyone is easy and worthless — a misleading subject produces an open, a moment of irritation, and sometimes a spam complaint.

Manage expectations before starting. Subject lines are the least consequential variable in cold outbound, well behind audience, offer and first line. They're the most tested because they're the easiest to test. If reply rates are the problem, the subject line is rarely the cause.

What works

Short. Three to five words. Mobile clients truncate around 35 to 40 characters, and most cold email is first seen on a phone.

Lowercase or sentence case. Title Case reads as a newsletter. quick question looks like a person typed it; Quick Question About Your Sales Process looks like a campaign.

Specific to them. The strongest subject lines reference the thing the email is actually about.

your Capterra reviews
SDR hiring
the Munich office
{{company}} + {{our company}}
question about your API docs

Curiosity without a trick. A subject that makes someone want to know more, and then the email delivers on it.

Plain. No punctuation marks, no emoji, no brackets, no "RE:" on a first email.

What doesn't

AvoidWhy
Quick Question!!!Exclamation marks read as promotional
RE: our conversationA lie, on a first email. People notice
🚀 Boost your revenue 300%Emoji plus a number plus a claim
Following up on my previous emailOnly valid if there was one
[Company Name] - Partnership OpportunityBrackets and title case, reads as bulk
Are you the right person to speak to about...Long, and answers itself
Free trial for {{company}}"Free" plus an offer, in a subject, from a stranger

The pattern in all of them: they read as something a marketer wrote to a list, and that's exactly what a filter and a recipient are both trying to detect.

Merge fields in subject lines

Use them sparingly and always with a fallback.

{{first_name}}, quick question becomes , quick question when the field is empty, which is both embarrassing and a strong bulk-mail signal. Set the fallback and test-render on 20 real leads before launching.

{{company}} is generally safer than {{first_name}} — company names are more reliably present and less likely to be malformed — but it also reads more like a mail merge. A specific fact usually beats either.

Matching the subject to the email

The subject sets an expectation. The first line has to meet it, or the open is wasted and the reader learns to skip your next one.

Subject: SDR hiring
First line: Saw you're hiring two SDRs in Munich — third sales role this quarter.

That works because it's continuous. Someone who opened because of the subject immediately sees why they were emailed.

Subject: quick question
First line: I hope this email finds you well. I wanted to reach out because...

That doesn't. The subject promised a question and the email opens with filler.

Preview text

The first line of the email shows next to the subject in most clients, effectively doubling the subject's length. Almost nobody uses it deliberately.

Because the first line of a good cold email is already the specific observation, this usually takes care of itself — as long as the email doesn't open with "Hi {{first_name}}," on its own line, which wastes the preview on a greeting.

Some clients let you set a hidden preheader. For plain-text cold email, don't bother. The first line is better used as the actual first line.

Follow-up subjects

For step two, threading onto the original email is usually right. It gives context and looks like a normal follow-up rather than a new campaign.

For steps three and four, a fresh subject gives a second chance at the open, particularly if the first was ignored rather than rejected.

What not to do: Re: Re: Re: quick question, or a subject that guilts. Did you see my email? reads as entitled, and the honest answer is usually "yes, and I didn't reply".

Testing

If you're going to test, do it properly or don't do it.

  • One variable. Subject only, everything else identical.
  • Several hundred sends per variant. At 100 each, a 48% versus 52% result is noise.
  • Split randomly, not by list order — list order correlates with source, region and company size.
  • Measure replies, not just opens. A subject that lifts opens and drops replies attracted the wrong people.
  • Run it once and move on. Subject-line testing has a low ceiling.

And accept the measurement problem: open tracking is unreliable. Apple Mail Privacy Protection pre-fetches images and inflates opens; image blocking suppresses them. Directional over time, not accurate in absolute terms. This is the main reason not to over-invest in subject-line optimisation — you can't measure it well.

When low open rates aren't a subject line problem

An open rate under about 25% on a cold list usually means filtering, not a weak subject. Check in this order:

  1. Authentication — SPF, DKIM, DMARC passing. See spf-dkim-dmarc-setup
  2. Domain reputation and warmup — see email-warmup
  3. List quality — bounces, wrong-fit, stale addresses
  4. Content triggers — links, images, attachments. See spam-filter-triggers
  5. Then the subject line

A sudden drop from 50% to 15% with no content change is a deliverability event, not a copy problem. Rewriting subject lines in response wastes a week.

Twenty that work

quick question                    {{company}} + {{us}}
your Capterra reviews             SDR hiring
the Munich office                 your API docs
question about {{company}}        outbound
re: your changelog                idea for {{company}}
worth 10 minutes?                 the checkout flow
your job post                     two SDRs
wrong person?                     {{company}}'s onboarding
one thing                         about your pricing page
saw your Series A                 not sure if you own this

Most are short, specific and unremarkable. That's the point — they look like internal email, which is the register you want.

Don't do these

  • Don't mislead. An open you tricked someone into is worse than no open.
  • Don't use "RE:" or "FW:" on a first email.
  • Don't put a claim or a percentage in the subject.
  • Don't use emoji in B2B cold outbound.
  • Don't write a subject longer than the phone will show.
  • Don't optimise subject lines when reply rate is the problem.
  • Don't trust open-rate differences on small samples.

Related skills

cold-email-copywriting, email-sequence-design, spam-filter-triggers, outbound-metrics-benchmarks, email-deliverability

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.