Why 95% of LinkedIn DMs Get Ignored
You know the ones. Someone connects with you, and 30 seconds later you get a message that reads like it was written by an algorithm that learned English from LinkedIn posts:
Here's why most LinkedIn DMs fail immediately:
- Zero specificity: "Hi [Name]" could be sent to anyone. You can tell they didn't read your profile.
- Lead-with-pitch: A selling message as the opening move gets deleted before you finish reading it.
- Vague benefit: "Help you grow" or "increase visibility" means nothing without context.
- Assumption of urgency: "This week?" You don't know them, and you have no reason to drop everything.
- No personality: It reads like it came from a template. Because it did.
- Missing "why now": They didn't mention what triggered the outreach. A recent post? A job change? Something they said?
The good news: most of your competitors are sending these exact DMs. That means the bar for "actually good" is embarrassingly low.
The Principles of a LinkedIn DM That Works
Before we get to templates, here are the non-negotiable rules:
1. Make it personal—specifically
Reference something from their profile, a recent post, or a mutual connection. Not "I noticed you work in marketing" (everyone works in marketing). More like: "I read your post about why TikTok creators should use LinkedIn—completely agree about the audience overlap angle."
2. Keep it short
LinkedIn DMs on mobile are painful. If your message is more than three sentences, they're already skimming. Aim for two sentences in the first message.
3. No pitch in message one
Your job in the first message is to get them to reply. That's it. Not to make a sale, book a call, or explain your value prop. Build trust first.
4. Offer something for free first
A thought, a resource, an intro, an insight. Make it cost you something (time) and them nothing. Then ask.
5. Use a trigger event
The best DMs reference something that just happened: a post, a job change, a company milestone, a comment they made. "Just saw" is more powerful than "I noticed in your profile."
6. Sound like a human
Write like you talk. Contractions are your friend. Emoji sparingly, but not zero. You're writing to a person, not a robot.
The 6 LinkedIn DM Templates That Get Replies
Template 1: The Connection Request Note
You're reaching out cold. You have 300 characters to convince them you're worth their time. No pitch, just a reason.
How to use AI to customize:
Paste their last 3 posts into Claude or ChatGPT and ask: "What themes does this person care about? What problem are they discussing?" Then reference the specific theme in your note, not just their job title.
Claude
Better for nuanced personalization. Give it context about the person and ask: "Write a connection note that sounds like a founder (not a salesperson) reaching out."
ChatGPT
Faster for generating multiple variations. Ask it to create 5 versions targeting different angles (shared frustration, opposite take, question, etc.).
Template 2: The Genuine Compliment (No Pitch)
You actually liked something they made. Your message is just to say so. Surprisingly rare on LinkedIn.
AI hack:
Paste the post you liked into Claude and ask: "Give me 3 specific insights from this that surprised me or changed my thinking." Pick one. Reference it in your message.
Template 3: The Referral From a Shared Connection
You have a mutual connection. Leverage it.
When to use:
Only if the mutual connection actually said "you should talk to this person." Cold name-dropping dilutes the effect.
Template 4: The Collaboration Offer
You have an idea that involves them. You're not asking for their time; you're offering to do something together.
AI's role:
Ask Claude to brainstorm collaboration ideas based on what you both do. "I create podcast episodes, they create educational content. Give me 5 collaboration ideas that play to both strengths."
Template 5: The Job Seeker DM
You're looking to hire or looking for a role. This is different. People expect it more, but you still need specificity.
Template 6: The Podcast/Newsletter Guest Pitch
You want them on your show/newsletter. This works because you're asking them to be the hero.
How to Personalize Templates (Without Sounding Like an AI)
The tension: you want to reach hundreds of people, but DMs that feel templated get deleted immediately. Here's how to use AI to bridge that gap.
The Input-to-Output Method
Step 1: Create the base template. Write one example DM that genuinely works. Make it yours.
Step 2: Extract the variables. What changes person to person? Their name, a specific post they wrote, their job title, a mutual connection, etc.
Step 3: Give AI context, not just blanks. Instead of asking for: "Hi [First Name], I read your post about [Topic]..."
Ask Claude: "Here's my DM template. For Sarah, who writes about creator monetization and just started a Substack, what post would I reference? Write 3 options."
Then pick the one that feels genuine and drop it in.
The Voice Matching Method
If the person you're reaching out to uses a lot of emojis, short sentences, or specific jargon, ask AI to adjust your message to match that style—not to copy them, but to be more relatable.
In Claude: "This person writes in a very casual, emoji-heavy style. Here's their last post: [paste]. Rewrite this DM to match that voice, but keep it authentic to me."
The Trigger Event Method (Most Effective)
Rather than sending the same template to 100 people, collect "trigger events" and use them to personalize. A trigger event is something that just happened:
- They posted in the last 48 hours
- They changed jobs (visible on their profile update)
- They were mentioned/tagged in something relevant
- Their company hit a milestone or raised funding
- They commented on something you posted
DMs sent within 48 hours of a trigger event get 3-4x better reply rates than random outreach. AI helps you spot and leverage these moments.
Sales Navigator
LinkedIn's official tool to find and track people. Filters by job change, company growth, engagement. Trigger events are built in—use them.
LinkedIn Automation Tools: When They Help, When They Get You Banned
Let's be clear: LinkedIn's terms of service don't allow automation. But people do it anyway. Here's what's out there and what the risk-benefit actually looks like.
When Automation Makes Sense
Automation is useful for volume work that's not direct outreach:
- Scheduling your own posts (LinkedIn allows this natively)
- Batch-creating follow-up reminders ("Follow up if no reply in 7 days")
- Organizing leads into sequences (manual outreach, not automated)
- Tracking who viewed your profile or engaged with posts
The tools that help here are lower-risk because they're not impersonating human behavior at scale.
When Automation Ruins Your Account
High-risk moves that get accounts restricted or banned:
- Sending 100+ DMs per day without real personalization
- Auto-accepting connections and immediately sending a pitch
- Using bots to comment on posts with spammy messages
- Scraping profiles en masse and emailing people outside LinkedIn
Taplio
LinkedIn DM scheduling and connection request templates. Lower risk because you write the messages and choose when they go out. Good for sequences and follow-ups.
Meet Alfred
Multi-channel outreach (LinkedIn, email, Twitter). You can automate finding contacts and managing sequences. Safer than LinkedIn-only bots because it's not purely LinkedIn automation.
Expandi
LinkedIn DM automation, connection adds, profile views. High risk—this is exactly what LinkedIn hates. If you use it, keep volume low (max 20-30 per day) and never fully automate messaging.
Dux-Soup
Profile views, connection adds, automated responses. Lighter-weight than Expandi. Still against ToS, but if you're going to automate, this is more conservative.
The Honest Take
If you're smart about it: use automation for organizing and tracking, not for sending. Write your own DMs, even if you use a tool to schedule them. Send fewer messages but make them count. LinkedIn's algorithm catches patterns—if 500 accounts get the same message from your bot, you're done.
The creators winning at LinkedIn right now aren't using bots to scale outreach. They're using AI to write better DMs and they're sending fewer, more targeted messages.
How to Reply (And Actually Keep the Conversation Going)
Your DM got a reply. Now what?
The Response Pattern
Different replies demand different strategies:
They replied but asked a question
Answer it fully. Don't deflect to a call. Show you're here to help, not to close a deal.
They replied with "Thanks, not interested"
Respect it. Don't follow up twice. One line: "No worries. If this becomes relevant later, you know where to find me."
They asked for a call
Great. But before you jump on: Ask one more qualifying question. "What would be most useful to cover?" This prevents 30-minute calls that go nowhere.
They're interested but uncertain
Send something small and useful first. A resource, an article, an intro. Build trust before asking for their time.
They're interested and ready to move forward
Now you pitch. Now you set the call. Now you close.
In other words: your first message gets the reply, your second message earns the call.
The Follow-Up That Doesn't Annoy People
They didn't reply. Do you follow up? When? How many times?
Rule 1: One follow-up, maximum. If they didn't reply to your first message, they saw it and chose not to engage. A second message doesn't make you more interesting; it makes you persistent.
Rule 2: Follow up with a hook, not a repeat. Don't send the same message again. Send something new that might change their mind.
Rule 3: Time it right. Follow up after 5-7 days, not 2 days. You want them to forget the first message so the second feels new.
Rule 4: Accept the L. Some people just aren't going to reply. That's okay. Keep your list of "warm contacts" and re-engage them later when they do something that's a fresh reason to reach out.
Creator-Specific DM Playbook: Brand Deals, Collabs, and Monetization
If you're a creator (newsletter writer, podcaster, YouTuber building on LinkedIn), your DMs are probably different. People aren't hiring you; they want to partner with you.
The Brand Partnership Inquiry
What they're asking: "Can you mention our product to your audience?"
What you should do: Reply with your media kit. If you don't have one, create a one-pager with:
- Your audience size and breakdown
- Engagement rate (not vanity numbers)
- Your rate (if you charge)
- What brands have worked with you before
- Two to three audience testimonials or case studies
Most creators skip the media kit and end up negotiating blindly. Have the numbers ready.
The Collaboration Ask
What they're asking: "Can we do something together?"
What you should ask back:
- What's the scope? (one guest post, ongoing series, one-off collab?)
- What's the timeline?
- What's in it for both of us? (audience size, credibility, revenue share?)
Vague collaborations kill momentum. Get specific before you commit.
The Monetization Question
What they're asking: "How do you make money from your newsletter/podcast?"
Your move: This is a good sign. They're interested in partnership or learning your model. Answer honestly. But don't over-explain or feel bad about making money. Monetized creators are credible creators.
Using ChatGPT vs. Claude for DM Personalization
ChatGPT (Free or $20/mo Plus)
ChatGPT is faster and better for generating multiple variations quickly. Use it when you want quantity and speed.
Best for: "Generate 5 versions of this DM approach—one as a question, one as a compliment, one as a referral, one as a collaboration offer, one as a funny take."
Prompt that works: "I'm reaching out to Sarah, who writes about creator monetization and just started a Substack. Draft 3 different DM approaches that would make her want to reply. Each should be under 100 words and feel human."
Claude (Free or $20/mo Pro)
Claude is better at nuance and matching tone/voice. Use it when you need your DM to sound like YOU, not like a bot.
Best for: "Make this DM sound less corporate and more like how I talk in DMs."
Prompt that works: "I want to reach out to Jordan about automating customer onboarding. Here's how I normally write. [paste 2-3 real DMs you've sent]. Now draft an outreach message that sounds like me, not like an AI."
Pro tip for both: Give the AI examples of good DMs and bad DMs, then ask it to "analyze what makes the good ones work" before drafting new ones. It'll generate better templates.
FAQ
LinkedIn doesn't publish exact limits, but patterns suggest 20-30 per day is safe if they're genuine messages. Once you hit 50+, you're entering shadowban territory. But here's the thing—it's not about volume, it's about quality. Five great DMs will get more replies than fifty templated ones. Focus on that.
Cold DMs work, but connection-first is safer. Send a connection request with a personalized note (see Template 1), wait for acceptance, then send your actual outreach. It takes two extra days but gets better reply rates and doesn't annoy people.
Send it immediately, no friction. These are qualified signals. They're actually interested. Make your pitch deck/media kit clear, professional, and to the point. One page is better than three. Use it to start a conversation, not close a sale.
Track: (1) open rate (did they read it?), (2) reply rate (did they respond?), (3) outcome rate (did it lead to something?). You'll see patterns. If one template consistently outperforms others, you've found something. If you're getting 30% reply rate on certain approaches and 5% on others, double down on what's working.
Go Deeper: Your DM Strategy Playbook
DMs are just one part of your LinkedIn strategy. To actually build an audience and convert them:
- Read the complete LinkedIn creator strategy guide—covers posting, growth, monetization.
- Optimize your profile so people want to DM you first.
- Master thought leadership writing so your DMs reference real value.
- Use AI to scale from 0 to 10K followers.