Most course creators personalize emails by inserting a first name into the greeting and calling it done. That's table stakes — it hasn't moved open rates in years. Real personalization means writing different emails for different people based on what they've actually done: which pages they visited, what they've purchased before, where they dropped off. Teams using this kind of AI-driven behavioral personalization report significantly higher reply rates than generic blasts.
What you’ll walk away with:
- Segment-specific email variations for different subscriber stages
- A personalization strategy beyond merge tags
- Automation rules delivering the right message to the right segment
Why "[First Name]" isn't personalization anymore
I've watched the data shift over the past few years. Merge fields were impressive in 2015. Now every email platform does them by default, and subscribers barely register the difference. What actually moves engagement is relevance — writing to someone's specific situation rather than addressing them by name while saying the same thing you'd say to everyone else.
Here's a real example: one course creator I worked with was sending the same launch email to their full list. Open rate: 12%. We helped them segment into three groups — new subscribers, past students, and sales-page visitors who hadn't enrolled — and used ChatGPT to write a different version for each. Open rate jumped to 34%. Same offer, same course, same week. The only difference was that each email spoke to where that person actually was.
With most emails now opened on mobile, this is even more critical. You have roughly two seconds of attention on a phone screen. A subject line that references something the subscriber actually did ("You looked at the yoga retreat curriculum — here's what students said about it") earns those two seconds. A generic "New course available!" gets swiped away.
How to segment before you write
In Kit, create segments based on subscriber behavior. The most useful segments for course marketing aren't demographic — they're behavioral:
- New subscribers (joined in the last 30 days) — they need orientation, not a hard sell
- Sales-page visitors (clicked a link to your course page but didn't buy) — they're interested but have objections
- Past students (purchased a previous course) — they already trust you; the pitch is different
- Cold subscribers (haven't opened in 90+ days) — re-engagement, not enrollment
You don't need all four on day one. Start with the first three. Each one gets a different email — different subject line, different opening, different emphasis. That's what ChatGPT helps you produce without spending a full day writing.
Writing segment-specific drafts with ChatGPT
The key is giving ChatGPT enough context about each segment to generate copy that actually diverges. If you just say "personalize this email," you'll get the same email with slightly different adjectives. Instead, describe the segment's situation and intent.
For subject lines specifically, keep them between 28 and 50 characters. That's the range that performs best on mobile in 2026 — long enough to be specific, short enough to display fully on a phone screen. ChatGPT tends to write subject lines that are too long unless you set the character constraint explicitly in your prompt.
Prompts to try
I teach [course topic] to [audience]. I'm launching a new course called [name] on [date]. Price is [price]. Write three versions of a launch email for these segments: 1. NEW SUBSCRIBERS (joined in the last 30 days, have never purchased). They need to understand what the course is and why it matters. 2. SALES-PAGE VISITORS (clicked through to the course page but didn't enroll). They're interested but hesitating — address likely objections. 3. PAST STUDENTS (bought a previous course from me). They already trust me — emphasize what's new and offer a returning-student incentive. For each version: write a subject line (28-50 characters), a preview text line, and the full email body (150-200 words). Tone: warm and direct, like writing to one person. No hype. No "limited spots."
This prompt works because it forces ChatGPT to write from three different starting points. The new-subscriber version will explain; the past-student version will build on existing trust. That's the kind of variation that matters.
Here's an email I'm sending to promote my course: [paste your draft] Rewrite the subject line and opening paragraph for someone who: - Visited my sales page yesterday but didn't enroll - Has been on my email list for 6+ months - Has opened 4 of my last 5 emails This person is warm but hasn't committed. The email should acknowledge their interest without being creepy about tracking. Subject line must be under 45 characters.
This works well when you already have a decent email draft and want to create a variant for your most engaged non-buyers — the segment most likely to convert with the right nudge.
Generate 10 subject line variations for a course launch email. Course: [name and one-sentence description] Audience: [who they are] Five should use a curiosity angle (make them want to open to learn more). Five should use a specificity angle (tell them exactly what they'll get). All must be 28-50 characters. No exclamation marks. No "you won't believe" or "don't miss" patterns.
Splitting by angle prevents ChatGPT from generating ten variations of the same idea. You'll quickly see whether curiosity or specificity resonates more with your particular audience — and you'll have testable options for both.
The human layer
AI-personalized email is powerful, but it can feel hollow if the underlying relationship isn't real. The higher reply rates I mentioned earlier come from teams that combine AI-generated copy with real subscriber data — real interactions, real behavior patterns, real course outcomes. The AI handles the writing scale. The human handles the truth.
Here's what I'd recommend: after ChatGPT generates your segment-specific drafts, read each one and ask yourself, "Would I send this to [specific person] in this segment?" If you can picture a real subscriber and the email feels right for them, it's working. If it feels like a clever algorithm wrote it, it needs your voice layered back in. The best personalized emails read like they were written by someone who pays attention — because they were.
What it gets wrong
ChatGPT's biggest weakness with personalized email is that it fakes specificity. It'll write "I noticed you've been exploring our resources" when it has no idea what the subscriber actually did. You need to replace these vague references with real behavioral data from Kit — what they clicked, what they bought, when they joined. The prompt can describe the segment, but the final copy should reference actual behavior, not AI-generated guesses about behavior.
It also struggles with tone calibration across segments. The past-student version often sounds too similar to the new-subscriber version because ChatGPT defaults to explaining things. Past students don't need the basics — they need to know what's new and why it matters to them. You'll typically need to edit the past-student variant more heavily than the others to match the familiarity those subscribers expect.
And it will overuse personalization tokens if you let it. An email that mentions the subscriber's name four times and references their behavior in every paragraph feels surveillance-like, not personal. Once or twice is enough. Let the relevance of the content do the personalizing, not the name-dropping.
From drafts to delivered
Once your segment-specific emails are polished, the delivery side is straightforward. Kit's automation features let you set up rules that send the right version to the right segment automatically. On Ruzuku, your enrollment page handles the rest — when a personalized email earns the click, the course page makes enrollment simple, with payment and access in one place.
Related guides
- How to Segment Your Audience Using Kit — set up the behavioral segments that make personalization possible
- How to Use AI to A/B Test Email Subject Lines — test which personalized subject lines actually perform best
- How to Build a Course Launch Email Sequence Using Kit — the full sequence framework these personalized emails fit into
- How to Write Your Course Launch Email Sequence — the complete strategy guide for email-driven course sales