AI DM/Outreach Writer — Cold intro, partnership, sales, follow-up

Generate 3 DMs with 3 different angles (problem-led, compliment-led, mutual-led) for cold outreach on LinkedIn / Instagram / Twitter / Threads. 6 purposes, tone matched to platform.

3 angles6 purposes4 platformsAnti-spam

About you (sender)About recipient

DM purpose

Platform

Why use this tool

🎯

3 distinct angles

Problem-led (mention pain point), Compliment-led (specific, not generic), Mutual-led (common ground). A/B test which converts best.

🚫

Avoid generic templates

AI refuses opening lines like 'Hi, I'm very impressed with your profile'. Each DM mentions one specific detail about the recipient to prove it's not mass-sent.

🌐

4 platforms, different contexts

LinkedIn formal ≤500c. Instagram casual ≤300c with emoji. Twitter ≤250c light slang. Threads ≤280c conversational.

How to use

1Describe yourself (sender): role, company, niche, goal.
2Describe recipient: name, role, company, niche / why you want to reach.
3Pick purpose: Intro / Partnership / Sales / Follow-up / Networking / Reconnect.
4Pick platform: LinkedIn / Instagram / Twitter / Threads.
5Generate → 3 DMs with 3 different angles. Pick a favorite, personalize, send.

DM outreach — why 99% do it wrong

Cold DM reply rate is typically <2% because senders use generic templates. 'Hi, I'm impressed with your profile. I'd like to connect...' — recipients get 50 similar messages a week, scroll past.

Effective DMs differ: (1) Mention SPECIFICALLY about recipient (recent post, achievement, niche), (2) Value for recipient (not just an ask), (3) Soft ask or no ask (just relationship-build).

3 most effective angles: Problem-led ('I noticed [pain point] in your industry — I've solved similar...'), Compliment-led (SPECIFIC — 'Last week's thread on Y was great, especially point Z'), Mutual-led ('We both [niche], both know Z, both attended event W').

This tool builds DMs with 3 angles for the same sender-recipient. A/B test which angle converts. Each angle follows HOOK + VALUE + ASK structure.

✓3 DM variations / 3 angles (problem, compliment, mutual)
✓6 purposes: intro, partnership, sales, follow-up, networking, reconnect
✓4 platforms: LinkedIn, Instagram, Twitter, Threads
✓Auto char limit: LinkedIn ≤500, IG ≤300, Twitter ≤250, Threads ≤280
✓AI mentions SPECIFICALLY about recipient (not generic)
✓Avoids stilted formality, no CAPS LOCK
✓JSON output, validated
✓Rate limit 20/hour/IP

Use cases

Cold B2B sales

Pitch service to CMO/CEO — Problem-led DM mentioning pain point they shared.

Partnership outreach

Propose podcast collab, content trade — Mutual-led DM noting common niche.

Event follow-up

Met 10 people at event → personalized follow-up DM per person.

Investor outreach

Founder → investor — Compliment-led mentioning their recent investment thesis.

Influencer collab

Brand → creator — Problem-led 'I noticed you struggle with content burnout — we have a solution...'.

Reconnect old contact

Met a year ago, lost touch → Mutual-led casual reconnect, non-transactional.

Tech

Backend /api/ai/dm POST {sender, recipient, purpose, platform}. Validate sender + recipient ≤500 chars.

Non-streaming JSON output. Temp 0.75. Prompt emphasizes 'NO generic templates' + 'mention specific details from recipient context'.

3 angles hardcoded in prompt: Problem-led (variant 1), Compliment-led (variant 2), Mutual-led (variant 3) — ensures diversity.

FAQ

Does AI 'mention specifically' correctly?

AI uses info you provide in 'recipient context'. More detail = more specific mention. If only 'CEO marketing', mention will be generic.

Can DMs be in English?

Default VN (anchored). For pure EN: paste sender + recipient in English, AI outputs English DM. Or use /en/viet-dm.

What's the realistic reply rate?

Depends on many factors (niche, recipient activity, channel). This tool boosts % vs generic templates, but still personalize edits before sending.

Spam policy concerns?

Tool doesn't mass-send. User copies one and sends manually — full control. Still, avoid >20 DMs/day per account to avoid flags.