Cold DM That Actually Works Prompt
Build a research-backed 3-message cold outreach sequence with personalization framework, variations, and a self-scoring rubric.
Use This When
General business and marketing workflows.
Inputs Needed
Goal, context, audience, constraints, examples, desired output, deadline.
Expected Output
Clear structured answer with assumptions, recommendations, examples, and next steps.
The Workflow Prompt
You are a senior consultant. Objective: Cold DM That Actually Works Prompt Context: Build a research-backed 3-message cold outreach sequence with personalization framework, variations, and a self-scoring rubric. Original task: You are an outbound sales expert who has written cold messages that generated over $2M in pipeline revenue. You never send generic messages. Every DM you write feels handcrafted because it references something specific about the recipient.I sell [YOUR PRODUCT/SERVICE] at [PRICE POINT]. My ideal prospect is [DESCRIBE THEM]. I reach out on [LINKEDIN / INSTAGRAM / X / EMAIL].Before writing a single word, create a prospect research checklist. List the 7 things I should look up about every prospect before messaging them. Include their recent content, company news, mutual connections, tech stack, and public pain points. For each item, tell me exactly where to find it and what to look for. This research makes the message feel personal instead of spammy.Write the first message in a 3-message sequence. This message should NOT pitch anything. Its only goal is to start a real conversation. Open with a specific observation about their business. Reference something they posted, built, or said recently. Ask one thoughtful question that shows you understand their world. Keep it under 50 words. No generic compliments. Nothing that sounds mass-sent.Write message 2, sent 3 days later if no response. Add value without asking for anything. Share a specific insight or resource relevant to their situation. Keep it under 40 words. Then write message 3, sent 5 days after message 2. This is the soft pitch. Connect your offer to the specific problem you identified. Include one clear next step. Under 60 words. End with a question.Generate 3 complete variations of this sequence for different prospect types in my niche. Then list 10 rules I should never break when sending cold DMs. Include the exact words and phrases that kill response rates. End with a scoring rubric to grade my own DMs before sending: rate personalization, value, clarity, and call-to-action on a 1-5 scale. Inputs I may provide: Goal, context, audience, constraints, examples, desired output, deadline. Operating instructions: - First, restate the objective in one clear sentence. - If critical information is missing, ask up to 5 focused questions. If there is enough information to proceed, make practical assumptions and label them. - Use a Detailed response style. - Be specific to the business, audience, channel, and constraints provided. - Avoid generic AI advice. Give concrete recommendations, examples, templates, copy, or steps I can use. - When current facts, competitors, laws, prices, policies, or market claims matter, use current research and cite sources. - Do not expose hidden chain-of-thought. Provide a concise rationale or decision summary instead. - End with a short QA checklist that helps me verify the output. Required output: Clear structured answer with assumptions, recommendations, examples, and next steps. Caution: Avoid generic output; require concrete examples, assumptions, and next steps.
QA Follow-Up Checklist
After the AI returns its output, verify against:
- Output is specific to the provided business/context.
- Assumptions are clearly labeled.
- No unsupported claims without source checks.
- Next actions are clear and usable.
Follow-Up Prompt
Now turn the result for 'Cold DM That Actually Works Prompt' into a client-ready version: tighten wording, remove fluff, add missing assumptions, and provide the next 3 actions.
Avoid / Cautions
Avoid generic output; require concrete examples, assumptions, and next steps.
How Different Verticals Use This Workflow
Restaurant & Hospitality
A hospitality tech founder selling to restaurant owners DMs 40 prospects on Instagram, opening each with a specific observation about their recent post (new menu launch, expansion, hiring). Reply rate hits 22% — books 8 first conversations and 3 demos from a single week of personal DM outreach.
Retail & E-commerce
A B2B service provider targeting DTC founders DMs 30 prospects on X with a specific observation about their recent thread. Reply rate hits 16% — books 5 first conversations and 1 paying client from the first 30-DM batch.
Professional Services & B2B
A fractional CMO DMs 25 founders on LinkedIn with a specific observation about a recent podcast appearance or LinkedIn post. Reply rate hits 28% — books 7 first conversations and signs 2 new retainer clients from the first 50-DM batch.
Beauty & Personal Care
A beauty brand consultant DMs 20 indie founders on Instagram with a specific observation about their recent launch or content. Reply rate hits 35% — books 9 first conversations and 3 paid engagements from a week of cold DM outreach.
Local & Trade Services
A trade-specific software company DMs 30 contractor business owners on Facebook (where this audience lives) with a specific observation about their recent project posts. Reply rate hits 18% — books 6 first conversations and 2 paid demos from the first batch.
Frequently Asked
What inputs make a cold DM actually get a reply vs get archived?
Three things: a specific observation about the prospect that proves you researched them (their recent post, a podcast they were on, a hire they just made), the platform-appropriate length (Instagram DM under 35 words, LinkedIn under 70, X under 50), and zero pitch in the first message — its only job is to start a conversation. The 'I'd love to chat about X' close is the kiss of death; cut it.
Should I use ChatGPT or Claude Sonnet for cold DMs?
Claude Sonnet 4.6 — its conversational defaults match DM tone better and it resists the 'sales DM' patterns harder. ChatGPT GPT-5.5 tends to write LinkedIn-formal copy that gets archived on Instagram or X. Either way, the research is the work — AI writes the 30-word message in 90 seconds; you spend 6 minutes researching the prospect first. The research is what makes the message land.
How is this different from a cold email sequence?
Cold DMs work on platforms where the recipient has voluntary attention (Instagram, X, LinkedIn). The bar is higher — they have to feel personal in 30 words. Cold emails work on inbox attention (broader reach, more tolerant of length). DMs convert higher per touch (10-15% reply vs 3-5% for cold email) but volume is lower because the research per message is heavier. Use both, not either.
When is a cold DM the wrong channel?
When your prospect is over 50 and B2B — they don't check DMs reliably. When your offer is over $20K — DMs feel too casual for the trust required at that price. And when you can't sustain personal research per prospect; DMs done at template scale (30 prospects a day with a template) damage your account and your brand. Cap at 8-12 deeply researched DMs per day.