ChatGPT for LinkedIn Outreach: Prompts + Templates That Work

Using ChatGPT for LinkedIn outreach messages can turn a slow, manual task into a fast, personalized one, if you prompt it well and edit what it gives you. LinkedIn rewards messages that feel like they came from a curious human who actually looked at your profile, and punishes anything that smells like a mass blast. That tension is exactly where AI helps: it can draft and personalize at speed, while you supply the judgment that keeps every message human. This guide gives you the prompts and templates that work, plus the mistakes that get accounts restricted.

We build and run outbound systems for B2B companies, and LinkedIn is one channel we orchestrate alongside email inside them. Use ChatGPT as a drafting assistant, not an autopilot, and LinkedIn becomes a powerful part of a multi-channel motion.

Why LinkedIn outreach is different from cold email

LinkedIn is a smaller, more personal space than the inbox. Messages are short, the reader can see your photo and profile instantly, and there is a visible social layer, you are a real person reaching out, not an anonymous email address. That changes the rules. Over-formal, pitch-heavy messages that might survive in email die on LinkedIn, where everything reads more like a conversation between two people.

It also has hard limits. Connection request notes are capped at around 300 characters, and LinkedIn actively restricts accounts that send too many requests or messages too fast. So the goal is not volume for its own sake, it is a small number of sharp, personalized touches. ChatGPT helps you make each one specific without spending ten minutes per prospect. This multi-channel thinking is core to our outbound service.

How ChatGPT helps with LinkedIn outreach

ChatGPT is strong at three LinkedIn jobs: reading a profile and pulling out a genuine personalization hook, drafting a short connection note or message in seconds, and generating variations so your outreach does not become a copy-paste template everyone recognizes. Give it the raw material, a pasted profile, an About section, a recent post, and it will turn that into something specific to the person.

What it will not do is understand your ideal customer, keep your account safe, or judge whether a message actually sounds like you. Those stay with you. Used inside those limits, it is a genuine speed multiplier for personalized social outreach.

Prompts for connection requests

Keep these under 300 characters and make the accept feel natural.

• "Write a LinkedIn connection request under 300 characters to a [job title] at a [industry] company. Reference this profile detail: [paste]. No pitch, just a genuine reason to connect."
• "Give me 5 connection note variations based on this recent post by the prospect: [paste]. Each under 300 characters, curious not salesy."
• "Write a connection request that mentions something we have in common based on our two profiles: [paste both]."
• "Rewrite this connection note to sound less like a sales template and more like a peer reaching out: [paste]."
• "Write a connection request that compliments a specific idea in their content without being fawning: [paste]."

Prompts for the first message

Once connected, the first message should open a conversation, not deliver a pitch.

• "Write a first LinkedIn message under 60 words to a new connection who is a [job title]. Reference why we connected: [paste]. End with a light, genuine question, no pitch."
• "Turn this profile detail into an opening message that shows I actually read their profile: [paste]."
• "Write 3 first-message variations that ask about the prospect's work instead of talking about mine."
• "Rewrite this message to remove anything that sounds like a sales script: [paste]."
• "Write a first message that references the prospect's recent post and invites their take: [paste]."
• "Give me an opener that offers a relevant resource instead of asking for a call."

Prompts for follow-ups

Most LinkedIn conversations need a nudge. Make each one add something.

• "Write a LinkedIn follow-up to someone who did not reply to my first message. Add a new, relevant insight and keep it under 50 words."
• "Write a value-first follow-up that shares a useful tip related to [topic] and asks nothing."
• "Give me three ways to re-open a LinkedIn conversation without saying 'just following up.'"
• "Write a gracious final message that closes the loop and leaves the door open."
• "Rewrite this follow-up so it sounds patient and helpful, not pushy: [paste]."

Prompts for personalization at scale

This is where AI saves the most time on high-volume social outreach.

• "I will paste 10 prospects' About sections. For each, write one connection note under 300 characters. Return them numbered."
• "Given these profile variables ({{role}}, {{company}}, {{recent_activity}}), write a message template where the personalized part is one specific sentence."
• "Cluster these 15 prospects by likely priority based on their profiles, and write one tailored opener per cluster: [paste]."
• "Rewrite this generic message to accept a {{custom_hook}} pulled from each profile."

Templates ChatGPT can adapt

Give ChatGPT these as starting structures and ask it to personalize each one.

Connection request (no pitch): "Hi {{first_name}}, I came across your post on {{topic}} and thought your point about {{specific_idea}} was spot on. Would be glad to connect and follow your work."

First message after accept: "Thanks for connecting, {{first_name}}. I work with {{audience}} on {{problem}}, but mostly I was curious how you are handling {{specific_challenge}} at {{company}} these days. Always interested in how different teams approach it."

Value-add follow-up: "{{first_name}}, unrelated to anything I want from you, this {{resource}} on {{topic}} made me think of your situation at {{company}}. Might be useful. No agenda."

Feed each template to ChatGPT with a real profile and ask it to make the variables specific. The structure stays human, the details make it personal.

Combine LinkedIn with email, do not silo it

The biggest mistake is treating LinkedIn as a standalone channel. Buyers do not live in one place, and a prospect who ignores three emails might reply to a well-timed LinkedIn message, or vice versa. The strongest outreach weaves the two into one orchestrated sequence: a LinkedIn connection warms up a cold email, an email references a shared LinkedIn interaction, and the touches reinforce each other. That orchestration across channels is exactly what compounds, and it is the backbone of our case studies.

Where LeadHaste fits

We orchestrate LinkedIn and email together inside one outbound system we build, launch, and run for you. AI helps us draft and personalize, but the system handles what matters most: verified targeting, safe sending practices, multi-channel sequencing, reply handling, and month-over-month optimization.

Everything we build, you own, including the infrastructure and the relationships. And we guarantee performance, so if we miss the targets we agree on, your billing pauses. Explore the full service or read more on our blog.

Ready to make LinkedIn part of a system that compounds?

ChatGPT can help you write better LinkedIn messages, but messages alone are not a pipeline. Orchestrating them with email into one system is. We build that system and prove the results before you pay.

Book your free pilot →