Skip to content
sinc-LLMBook a 30-minute call

Engineering instrument

Email Writing Prompt Template — AI-Structured

A six-field email prompt template for outreach, follow-ups and newsletters. Specify audience, tone, requirements and the intended action; review the draft and measure results separately.

Inputs, specification fields, and output format for the task described here.

For an email in a direct, specific voice, define the audience, tone, purpose and banned phrases before drafting. Compare the generated email with that brief and review subscriber engagement using your own records; the template does not establish an open rate.

x(t) = Σ x(nT) · sinc((t − nT) / T)
Email voice is a high-frequency signal. Without the full sinc structure, the model defaults to its statistical average — which is corporate and forgettable.

This template is for a cold outreach email from a B2B founder. The recipient psychology and CONSTRAINTS band do the heavy lifting — the model needs to understand both the sender's voice and the recipient's likely resistance:

{
  "formula": "x(t) = Σ x(nT) · sinc((t - nT) / T)",
  "T": "specification-axis",
  "fragments": [
    {
      "n": 0,
      "t": "PERSONA",
      "x": "You are a B2B founder who writes cold emails that get responses. Your email voice is confident but not arrogant — you respect the recipient's time above everything. You lead with value, not pedigree. Your emails are conversational, not transactional."
    },
    {
      "n": 1,
      "t": "CONTEXT",
      "x": "I'm reaching out to VPs of Engineering at Series B-C SaaS companies (50-300 engineers). I'm the founder of a developer productivity tool. The recipient likely gets 20-40 cold emails per day and has learned to spot and ignore the pattern. My tool saves engineering leads 4 hours/week on sprint reporting."
    },
    {
      "n": 2,
      "t": "DATA",
      "x": "Product: DevCycle. Key outcome: auto-generates sprint reports from Jira + GitHub activity. One specific customer win: CTO at a 120-person team said 'I got my Friday afternoons back.' Price: $29/engineer/month. Trial: 14 days, no credit card."
    },
    {
      "n": 3,
      "t": "CONSTRAINTS",
      "x": "No fake personalization ('I was researching your company and...'). No empty compliments. Do not open with 'I hope this email finds you well.' Do not use the phrase 'quick question.' Do not mention competitors. Do not use more than 2 sentences to introduce the product. The CTA must be a specific, low-commitment ask — not 'schedule a 30-minute demo.' Max 120 words total."
    },
    {
      "n": 4,
      "t": "FORMAT",
      "x": "Subject line (max 8 words, no emoji). Email body: 3-4 short paragraphs. CTA in final paragraph. No signature block — I'll add that. Plain text format, no markdown."
    },
    {
      "n": 5,
      "t": "TASK",
      "x": "Write a cold outreach email to a VP of Engineering introducing DevCycle and requesting a specific low-friction next step."
    }
  ]
}

Every banned phrase in the CONSTRAINTS band represents a pattern the recipient's brain has learned to pattern-match as "sales email" and mentally discard. "I hope this email finds you well" triggers skip. "Quick question" triggers skip. "I was researching your company" with no real personalization triggers skip.

The CONSTRAINTS band in an email prompt is essentially a list of psychological landmines you're defusing before the model writes a word. The model will avoid them actively, not accidentally include them. This is the primary mechanism by which structured email prompts outperform raw ones — not intelligence, but pre-elimination of the patterns that kill response rates.

Email prompt tip: Specify the word count as a maximum in CONSTRAINTS, not a target in FORMAT. "Max 120 words" is a hard constraint the model respects. "About 120 words" is a target the model ignores in favor of completeness. Email brevity is a constraint, not a preference.

Write a cold email to a VP of Engineering introducing my developer productivity tool DevCycle. It saves 4 hours/week on sprint reporting. Include a CTA. Keep it short.
CONSTRAINTS: No 'hope this finds you well.' No 'quick question.' Max 120 words. CTA = low-commitment ask only.
DATA: CTO quote 'I got my Friday afternoons back.' 14-day trial, no CC.
PERSONA: Confident, respects recipient time, leads with value.

The raw prompt produces an email that opens with "I hope this email finds you well" and closes with "Would you be open to a 30-minute call?" The structured prompt produces something a real VP of Engineering might actually read past the first sentence.

Try AI Transform — Structure Your Email Prompt Free
Scope and context

sinc-LLM / Email Writing Prompt Template