Day 12: Email Subject Line Lab
By 21 Days of AI · Last updated: July 4, 2026
The concept
The subject line is the gateway to the email. If it fails, the offer, design, copy, and CTA do not matter because the reader never sees them.
Marketers often write subject lines last, when the campaign is already built and attention has moved to scheduling. That is backwards. Subject lines deserve the same level of variation and testing as ad headlines, landing page headlines, or CTAs.
AI is valuable because it can generate a wide set of strategic variants quickly. The marketer's job is to evaluate relevance, risk, and fit.
Plain English
A good subject line earns the open without betraying the email inside.
Volume creates better options
Writing two subject lines produces a tiny solution space. Writing twenty forces you past the obvious.
The categories in today's prompt matter:
- Curiosity gap: creates interest without misleading
- Direct benefit: states the value clearly
- Specificity: uses concrete detail, numbers, or segment relevance
- Question: mirrors something the reader is wondering
- Urgency: gives a real time reason to act
- Contrarian: challenges a familiar assumption
Different campaigns need different strategies. A product update may need clarity. A thought leadership email may benefit from curiosity. A re-engagement email may need a direct or human tone.
Score before you test
AI's scoring is not data. It is pre-test review.
Use it to catch:
- subject lines that are intriguing but unclear
- subject lines that fit the offer but not the brand
- spam-risk language
- urgency that feels manufactured
- questions nobody is actually asking
- lines that overpromise what the email delivers
Then choose a primary and an A/B variant that test a meaningful difference. Testing two nearly identical lines rarely teaches much.
Preview text is part of the subject line
Subject line and preview text work together. The preview text should not repeat the subject. It should add context, tension, benefit, or reassurance.
Example pattern:
- Subject: A faster way to brief campaigns
- Preview: Use this before the creative review, not after.
The preview completes the thought and gives the reader a reason to open.
Build a subject line swipe file
Save your best subject lines by campaign type:
- launch
- nurture
- re-engagement
- newsletter
- event
- product update
- customer story
Record the segment, open rate, click rate, and notes. Over time, you will see what your audience opens and what only sounds clever internally.
Pair subject lines with intent
Different emails need different opening promises. A newsletter subject line can be curiosity-led because the relationship is ongoing. A password reset or onboarding email should be clear. A re-engagement email may need a more human tone. A launch email may need benefit and urgency. A customer education email may need usefulness.
Before choosing, identify the reader's likely state:
- busy and scanning
- actively evaluating
- newly interested
- cold or inactive
- already loyal
- skeptical
- expecting a resource
The same subject line can be strong for one state and wrong for another. AI's score is only useful after you define the context.
Avoid the cleverness trap
Subject lines can become too clever. They may get opens but create disappointment when the email does not deliver. That hurts trust and can damage future engagement.
Use this test:
If someone opens because of this subject line, will they feel the email delivered on the promise?
If the answer is no, revise. Open rate matters, but not in isolation. For most marketing emails, the goal is not an open. The goal is a meaningful action after the open.
Review deliverability basics
AI may suggest urgency language that creates risk. Watch for excessive punctuation, all caps, spam-trigger phrases, misleading "RE:" formats, or false scarcity. Deliverability is not only technical. Copy choices affect inbox placement and reader trust.
Ask AI to identify risk, but also use your email platform's data and deliverability guidance.
Today's practice
Choose one upcoming email. Run the prompt. Select:
- Primary subject line
- A/B test variant
- Preview text
- Reason for the test
Do not simply pick your favourite. Pick the pair that will teach you something about what your audience responds to.
Test meaningful differences
An A/B subject line test should compare different hypotheses, not tiny wording changes. "Save time with better reporting" against "Save more time with better reporting" teaches very little. "Save time with better reporting" against "Stop rebuilding reports every Friday" compares direct benefit with pain-specific language.
Before launching a test, write the hypothesis: I believe this variant will win because this audience is most motivated by this reason. After the test, record what happened. Over time, this becomes a subject line learning system, not a series of isolated guesses.
Read opens alongside clicks
Open rate alone can be misleading. A curiosity subject line may win the open and lose the click if the email does not satisfy the promise. A direct subject line may get fewer opens but stronger conversion because the right people self-select.
After the send, compare:
- open rate
- click rate
- click-to-open rate
- unsubscribes
- replies
- conversion
The best subject line is not always the one with the highest open rate. It is the one that attracts the right reader into the right action without damaging trust.
Segment-specific language matters
A subject line for executives should usually not sound like a subject line for practitioners. A subject line for new subscribers should not assume the same context as one for active customers. Segment language affects clarity and trust.
Ask AI to rewrite the top three subject lines for different segments:
- executive buyer
- hands-on practitioner
- inactive subscriber
- active customer
- trial user
This exercise often reveals which version is too broad. The more closely the subject line matches the reader's context, the less it has to rely on tricks to earn the open.
Before sending, read the line as the recipient. Would it feel relevant in their inbox today? If not, revise. The inbox is personal space, and relevance is what earns permission to keep showing up there.
One useful final check is to pair the subject line with the first sentence of the email. If the first sentence continues the promise smoothly, the open feels rewarded. If it changes topic, the subject line may be doing work the email cannot support.
Prompt of the day
Copy this into your AI tool and replace any bracketed placeholders.
Prompt
You are an email marketing specialist. Generate and evaluate subject line variants for an email campaign. Sender: [BRAND] Audience segment: [SEGMENT] Email purpose: [PURPOSE] Core offer or news: [WHAT THE EMAIL IS ABOUT] Desired action: [ACTION] Tone: [TONE] Generate 20 subject line variants across these strategies: 1. Curiosity gap, 3 variants 2. Direct benefit, 3 variants 3. Personalisation or specificity, 3 variants 4. Question, 3 variants 5. Urgency or scarcity, 3 variants where appropriate 6. Contrarian or surprising, 5 variants Score each from 1-10 for open potential, spam risk, and brand fit. Recommend the top 3 and explain why.
Your 15-minute task
Use an email you will send soon. Generate 20 subject lines, choose a primary and an A/B variant, and pair the winner with preview text.
Expected win
20 scored subject line options with a recommended primary line and test variant for a real email.
Power user tip
Ask for five preview text variants under 90 characters that add new information instead of repeating the subject line.
Finished today?
Mark this lesson done on this device. No account is required, and you can continue straight to the next day.
Want Day 13 in your inbox tomorrow morning?
Email delivery is optional. You can keep reading for free now, or use the starter sprint to get a short daily reminder.
Set up daily delivery