Email is still how most professional work gets done. And yet, most people spend way too long staring at a blank compose window, especially for messages that are high-stakes or emotionally tricky.
AI is genuinely good at email — but only when you give it the right context. The default "write me a follow-up email" produces something that reads like it was generated by a robot pretending to be a person. These five patterns solve that by providing the structure and specificity that produce emails you'd actually send.
The key insight: good email prompts need to capture the relationship context, not just the content. Who are you to this person? What's the history? What's the subtext? That's what these patterns account for.
One universal rule for all email prompts: always set a word count. Without one, AI writes emails that are 2-3x longer than they should be. The ideal professional email is under 150 words. The ideal cold email is under 100. Long emails don't get read — they get skimmed, misunderstood, or deleted.
Pattern 1: The Cold Outreach
This pattern generates first-contact emails that don't read like spam. It works for sales, networking, partnerships, or any situation where you're reaching out to someone who doesn't know you.
You are an experienced [your role — e.g., "business development manager"].
Write a cold outreach email with these details:
- Recipient: [Their name, title, company]
- What they care about: [Something specific — a recent post they wrote, a company initiative, a shared connection]
- What I'm offering: [Your value proposition in one sentence]
- The ask: [What you want them to do — reply, book a call, check out a resource]
Requirements:
- Subject line: under 8 words, no clickbait
- Email body: under 100 words
- First sentence must reference something specific about them (not generic flattery)
- One clear CTA at the end
- No "I hope this email finds you well" or similar filler
- Tone: confident and direct, not salesy or desperateWhy it works: The 100-word limit forces conciseness — most cold emails fail because they're too long. Requiring a specific reference to the recipient prevents generic templates. Banning filler phrases removes the most common AI writing crutches.
Example output snippet:
Subject: Quick question about your workflow automation
>
Hi Sarah,
>
Your LinkedIn post about cutting manual reporting time by 60% caught my attention — we've been solving the same problem for ops teams at mid-size SaaS companies.
>
We built an integration that connects Salesforce directly to your reporting dashboards, eliminating the CSV export step your team mentioned. Three companies in your space are using it.
>
Worth a 15-minute call this week to see if it fits?
When to adapt this pattern: For networking (not sales), replace "what I'm offering" with "what we have in common" and change the CTA to a coffee chat or virtual intro. For partnership outreach, add a "mutual benefit" field that explains what's in it for both sides.
Pattern 2: The Strategic Follow-Up
Following up without being annoying is an art. This pattern generates follow-ups that reference the previous interaction, add new value, and make it easy to respond.
You are writing a follow-up email on my behalf.
Context:
- Previous interaction: [What happened — meeting, call, email, event]
- When: [How long ago]
- Key point from last interaction: [The most important thing discussed]
- What I promised: [Any commitments I made — sending a resource, checking on something]
- What I want next: [The next step I'm proposing]
Write a follow-up that:
1. Opens by referencing a specific detail from our last interaction (not "just following up")
2. Delivers on any promise I made (or acknowledges it)
3. Adds one piece of new value (an insight, article, or relevant update)
4. Closes with a specific, low-friction next step
5. Total length: under 120 words
6. Tone: warm but professional, not pushyWhy it works: Referencing a specific detail signals that you actually paid attention. Adding new value gives the recipient a reason to engage beyond obligation. The low-friction next step (like "does Thursday at 2pm work?" vs "let me know when you're free") makes it easy to say yes.
Example output snippet:
Hi Mark,
>
Great conversation last Tuesday about the onboarding bottleneck your team is facing — especially the point about new hires taking 3 weeks before they handle tickets solo.
>
I pulled together the case study I mentioned about how Acme Corp cut their ramp-up time to 8 days. Attached.
>
Also noticed Gartner just published new benchmarks for SaaS onboarding efficiency — thought it might be useful context for your Q2 planning.
>
Would a 20-minute call Thursday afternoon make sense to dig into this?
When to adapt this pattern: If the previous interaction was negative or contentious, add "acknowledge any tension without reopening the debate" to the requirements. For very warm relationships, relax the word count to 150 and allow a more casual opening.
Pattern 3: The Difficult Conversation
Some emails are hard to write not because they're complex, but because the stakes are personal — delivering bad news, setting boundaries, addressing a conflict. This pattern handles tone carefully.
Help me write a difficult email.
Situation: [Describe what happened and why the email is necessary]
Recipient: [Who they are and your relationship to them]
My goal: [What outcome I want from this email]
What I want to avoid: [Specific tone or content to steer clear of]
Write the email following these guidelines:
- Lead with empathy or acknowledgment, not the bad news
- State the core message clearly in the middle — don't bury it
- End with a constructive path forward or next step
- Keep it under 150 words
- Tone: [Choose: firm but respectful / empathetic but direct / professional and neutral]
- Do not use passive-aggressive language
- Do not over-apologize — one acknowledgment is enoughWhy it works: The structure (empathy, then message, then path forward) mirrors best practices for difficult conversations. Specifying what to avoid prevents the AI from defaulting to either doormat mode (over-apologizing) or robot mode (cold and detached). The word limit prevents rambling, which is common when people are uncomfortable.
Example output snippet:
Hi Alex,
>
I appreciate the effort you put into the Q1 campaign proposal — the creative concepts were strong.
>
After reviewing the budget projections with finance, we're not able to move forward with the full scope. The numbers require us to cut the paid media portion, which reduces the plan to organic channels only.
>
I'd like to set up 30 minutes this week to rework the timeline around the organic strategy. I think there's still a strong campaign here — just scoped differently. Does Wednesday work?
When to adapt this pattern: For delivering truly bad news (layoffs, major project cancellation), remove the path-forward section if a path forward doesn't exist yet — false optimism is worse than honest uncertainty. For peer conflicts, add "avoid implying fault — focus on the situation, not the person."
Pattern 4: The Internal Update
Status updates and internal communications need to be clear, scannable, and respectful of people's time. This pattern creates structured updates that busy colleagues will actually read.
Write an internal update email for my team/stakeholders.
Context:
- Project/topic: [What this is about]
- Audience: [Who's receiving this — team, leadership, cross-functional group]
- Key update: [The main news in one sentence]
- Supporting details: [2-3 important specifics]
- Blockers or risks: [Anything the audience needs to know about]
- Action needed from recipients: [What you need them to do, if anything]
Format the email as:
- Subject line: [Project name] — [Status: On Track / At Risk / Update]
- TL;DR: One sentence summary at the top
- Body: Bullet points grouped under 2-3 clear headings
- Action items: Bold the names of people who need to do something
- Total length: under 200 words
- Tone: clear and factual, not overly formalWhy it works: The TL;DR respects busy readers who may only skim. Bullet points under headings make it scannable. Bolding action-item owners makes responsibilities unambiguous. The subject line format lets recipients prioritize without opening the email.
Example output snippet:
Subject: Project Atlas — On Track (Week 6 Update)
>
TL;DR: Backend API is complete, frontend is 80% done, on track for April 25 launch.
>
What's done:
- API endpoints shipped and passing all integration tests
- User authentication flow complete, including SSO
>
In progress:
- Dashboard UI — targeting completion by April 18
- Performance testing starts April 19
>
Needs attention:
- @David: We need final copy for the onboarding screens by Friday
- @Lisa: Legal review of the updated ToS is blocking the settings page
When to adapt this pattern: For "at risk" project updates, add a "proposed mitigation" section after the blockers. For leadership audiences who want less detail, reduce the body to a TL;DR plus three bullet points max. For cross-functional updates, add a "what this means for your team" section.
Pattern 5: The Customer Response
Responding to customers — especially unhappy ones — requires balancing empathy with efficiency. This pattern generates responses that acknowledge the issue, explain what's happening, and set clear expectations.
Write a customer response email.
Context:
- Customer's issue: [What they wrote about or complained about]
- What actually happened: [The real explanation]
- What we're doing about it: [Current status of the fix or resolution]
- Timeline: [When they can expect resolution]
- Compensation offered: [If any — discount, credit, free month, etc.]
Requirements:
- Open by acknowledging their specific frustration (not a generic "we're sorry")
- Explain what happened in plain language — no jargon or blame-shifting
- State the resolution and timeline clearly
- If we made a mistake, own it in one sentence
- Close with a specific person or channel they can reach if the issue continues
- Under 150 words
- Tone: empathetic and professional, not defensive or roboticWhy it works: Acknowledging the specific frustration (not just "we apologize for any inconvenience") shows you actually read their message. Owning mistakes in one sentence prevents both under-apologizing and over-apologizing. A specific contact for escalation gives them confidence the issue won't disappear into a void.
When to adapt this pattern: For angry customers who are threatening to churn, add "acknowledge the severity of their frustration" and include a retention offer. For positive feedback or feature requests, swap the structure entirely — lead with gratitude, share what's on the roadmap, and invite them to a beta if applicable.
Example output snippet:
Hi Jordan,
>
You're right to be frustrated — being charged twice for the same subscription is unacceptable, and I understand the urgency.
>
Here's what happened: a billing system error on our end processed your March payment twice on the 15th. This was our mistake.
>
I've already initiated a refund for $49.99. It should appear in your account within 3-5 business days. I've also flagged your account to ensure this doesn't recur.
>
If the refund doesn't appear by Friday, email me directly at support@company.com and I'll escalate it immediately.
Quick Tips for Email Prompts
- Include the relationship context. "My boss" vs "a peer" vs "a client I've never met" changes everything about tone and directness. A follow-up to your CEO needs a different structure than one to a colleague you've worked with for five years.
- Specify what NOT to say. AI defaults to certain email cliches. If you hate "I hope this finds you well" or "please don't hesitate to reach out," ban them explicitly. Build a personal "banned phrases" list and include it in every email prompt.
- Set a word count. Without it, the AI writes emails that are 2-3x longer than they should be. Most professional emails should be under 150 words. Cold emails under 100. Internal updates under 200.
- Provide the emotional subtext. "I'm annoyed but need to stay professional" or "I want to say no without damaging the relationship" gives the AI the nuance it needs to calibrate tone correctly.
- Iterate on tone, not content. If the facts are right but the vibe is wrong, say "same content but make it warmer" or "more direct, less hedging." It's faster than rewriting the whole prompt.
- Generate subject lines separately. After generating the email, run a follow-up prompt: "Generate 5 subject line variations for this email — under 8 words each." The subject line determines whether your email gets opened, so it deserves its own optimization pass.
- Include the email thread if replying. When responding to an existing conversation, paste the relevant previous messages. The AI can match the established tone and reference specific points from the thread.
When to Use Templates vs. Write From Scratch
Use these patterns when:
- You write the same type of email regularly (weekly updates, customer responses, outreach)
- You're stuck on how to start — the structure gets you past the blank page
- English isn't your first language and you want professional phrasing
Write from scratch when:
- The email is highly personal or emotionally sensitive in a way that's hard to template
- The situation has so much unique context that filling in the template takes longer than just writing
- You're replying in a thread with existing tone and context the AI doesn't have
Combining Patterns in a Workflow
Many email situations require more than one pattern. A common sequence:
- Cold Outreach (Pattern 1) to initiate contact
- Strategic Follow-Up (Pattern 2) after they don't respond to the first email
- Follow-Up with value add (Pattern 2 again, with fresh value) if there's still no response
For project communication, you might use:
- Internal Update (Pattern 4) for weekly stakeholder emails
- Difficult Conversation (Pattern 3) when a project hits a blocker
- Customer Response (Pattern 5) when the blocker affects a client
Each pattern references the same underlying situation but adapts the structure and tone for the specific communication need. Save your context once and reuse it across multiple patterns.
For emails you send repeatedly, SurePrompts' Template Builder lets you save and customize these patterns so each variant is a quick edit, not a rebuild from scratch.