# LinkedIn connection request message: 10 templates that get accepted, sent with your AI agent

Source: https://finchreach.com/blog/linkedin-connection-request-message
Published 2026-10-30 by Finchreach team. Tags: linkedin, invitations, how-to, mcp, campaigns.

Copyable LinkedIn connection request notes for 10 situations, follow-ups after they accept, when to withdraw pending invites, and how an AI agent drafts them.

## The short answer

A good LinkedIn connection request message is short, specific and about them. Say where you know them from (or what of theirs you read), give one honest reason to connect, and stop. No pitch, no calendar link, no paragraph about your company. Keep it well under LinkedIn’s limit for notes, which is up to 300 characters (LinkedIn has given some free accounts a shorter limit, so check what your note box allows); two sentences is plenty.

The basic shape:
```text
Hi {first_name}, [how you know them or what you read]. [One reason to connect]. Would be glad to connect.
```

Below are 10 copyable notes for common situations, follow-up messages for after they accept, how to clean up pending invitations that never got an answer, and how to have an AI agent draft all of it for you to approve.

## Why the note matters, and when to skip it

People decide on an invitation quickly: your name, headline and photo, and the note if there is one. A note that answers “why is this person asking?” removes the guesswork. A note that opens with a pitch answers it too, in a way that invites Ignore.

Sending no note at all is fine, and often the better choice, when:

- You just met in person or on a call, and they expect your request.
- You are clearly peers in the same small field, and your headline already explains it.
- You would only be writing filler (“I’d like to add you to my network”). An empty note beats a generic one.

LinkedIn also limits how many invitations you can send and how many personalized notes you can add, and the limits are tighter on free accounts. The exact numbers change over time and vary by account, so check what your own account shows rather than relying on a figure from a blog post. Either way, the limit is a good reason to spend notes on the people where context actually helps.

## Five principles for a note that gets accepted

1. Be specific. Name the event, the post, the school or the team. “Saw your talk” is weaker than “Your talk on on-call rotations at SRECon”.
2. Give a mutual reason. Why would connecting be useful or pleasant for both of you? Shared work, a shared topic, a shared community.
3. Don’t pitch in the invite. The invitation asks for a connection, not a meeting. Selling comes much later, if at all, and only once there is a conversation.
4. Write plainly. Short sentences, no buzzwords, no flattery you wouldn’t say out loud. Read it as the recipient would.
5. Personalize from what they shared. Their headline, their recent posts, a project on their profile. One real detail is worth more than three adjectives.

To find that detail, spend a minute on their profile and recent activity. A post from the last few weeks is usually the best opener: it is fresh, they chose to share it, and mentioning it shows you read it.

## 10 LinkedIn connection request templates

Every note below is under 300 characters as written. Replace the parts in braces, like {first_name} or {event}, then read it once more: if a detail doesn’t fit the person, cut it rather than force it.

1. You met at an event:
```text
Hi {first_name}, good talking with you at {event} about {topic}. I’d like to stay in touch and keep comparing notes on it. Would be glad to connect.
```

2. You read their post:
```text
Hi {first_name}, your post on {topic} was useful, especially the point about {specific_point}. I work on something similar at {my_company} and would like to follow your work. Glad to connect.
```

3. Same school or former company:
```text
Hi {first_name}, fellow {school_or_company} alum here. I saw you’re now working on {area}, which is close to what I do. Would be good to be connected.
```

4. Recruiter to candidate:
```text
Hi {first_name}, I recruit for {team} at {company} and your work on {project} stood out. No pressure at all, but I’d like to connect and share a role if the timing is ever right.
```

5. Candidate to hiring manager:
```text
Hi {first_name}, I applied for the {role} role on your team and I’m genuinely interested in {team_focus}. I’d like to connect and follow what your team is building.
```

6. Founder to a potential customer (no pitch):
```text
Hi {first_name}, I’m building in the {field} space and keep reading your posts on {topic}. I’d value being connected with people who know this work well.
```

7. Peer in the same role:
```text
Hi {first_name}, I’m also a {role}, at {my_company}. I liked how you described {topic} in your profile. Always good to know others doing the same job. Glad to connect.
```

8. After a webinar:
```text
Hi {first_name}, I attended {webinar} and your answer about {topic} stuck with me. Would be glad to connect and keep learning from what you share.
```

9. A podcast guest:
```text
Hi {first_name}, just listened to your episode on {podcast}. What you said about {topic} matched what I’m seeing at {my_company}. Would be glad to connect.
```

10. Reconnecting with someone you know:
```text
Hi {first_name}, it’s {my_name}, we worked together on {project} at {company} a few years back. Saw you’re now at {new_company}, congrats. Would be good to reconnect here.
```

> Recruiters and founders: the template is the invitation only. Mention a role or your product after they accept, in a message, and only if it is relevant to them.

## LinkedIn follow-up messages after they accept

Acceptance is not a yes to a meeting. A good first message continues the reason you gave in the invitation, and asks one easy question, or none.

Thanks, with a question:
```text
Thanks for connecting, {first_name}. Curious: how is your team handling {topic} these days? I’ve seen a few different approaches and I’m always interested in what works in practice.
```

Thanks, with something useful:
```text
Thanks for connecting, {first_name}. Since you wrote about {topic}, this {article_or_resource} might be useful: {link}. No reply needed.
```

One gentle follow-up:
```text
Hi {first_name}, following up once on my note about {topic}. If it’s not a priority right now, no problem at all.
```

When to stop: one follow-up after a week or so is reasonable. If there is still no answer, stop. Silence is an answer, and a third message rarely changes it. Keep engaging with their posts if you are genuinely interested, and let them come to you.

## Review and withdraw stale pending invitations

Invitations that sit unanswered for weeks are worth cleaning up: withdrawing old ones keeps your sent list focused on people who might still respond. On LinkedIn, they are under My Network, in the sent invitations view, each with a Withdraw button.

How long to wait is a judgement call. A few weeks is usually enough for someone to have seen the request; if they haven’t acted by then, they probably won’t. Keep in mind that LinkedIn may not let you re-invite the same person for a while after you withdraw, so don’t withdraw just to send again the next day.

With Finchreach connected, your agent can do the review: `linkedin_get_sent_invitations` lists your pending sent invitations with the invitee’s name, headline, profile, how long ago you sent it and the note. Ask it to pick out the stale ones, and it proposes `linkedin_withdraw_invitation` for each. Withdrawing is a write, so each one waits in your [approval queue](https://finchreach.com/approvals) until you approve it.

> Ask your assistant: “List my pending sent LinkedIn invitations. Show the ones older than four weeks in a table with name, headline and how long ago I sent them, and propose withdrawing those. I’ll approve them myself.”

## Doing it with your AI agent

Finchreach is a remote LinkedIn MCP server at `https://mcp.finchreach.com/mcp`. MCP (Model Context Protocol) is a standard plug that lets AI apps use outside tools. Once your assistant is connected, by OAuth sign-in in clients like Claude, ChatGPT and Gemini, or with an API key in clients that send headers, it can read your LinkedIn and draft invitations for you. Setup for each client is in the [docs](https://finchreach.com/docs).

For one person, the flow is simple:

1. The agent reads context: `linkedin_get_person_profile` for their profile card (name, headline, location, connection state), `linkedin_search` (a posts search, by keywords such as their name or a topic) to find what they have written, and `linkedin_get_feed` if they show up in your feed.
2. It drafts a note from that context, following the principles above.
3. `linkedin_send_invitation` sends the invitation, with the note optional and capped at 300 characters. It doesn’t go to LinkedIn yet: it waits in your approval queue, where you can edit the note, approve or reject. Nothing is sent until you approve.

> Ask your assistant: “Look at https://www.linkedin.com/in/example-person/ and their recent posts. Draft a LinkedIn connection note under 250 characters that mentions one specific thing they wrote, with no pitch. Then queue the invitation so I can review it.”

Note that viewing a profile with `linkedin_get_person_profile` shows up for them as a profile view, like visiting it yourself.

## Many people: use a campaign

For a list of people, ask your agent to build a campaign with `linkedin_create_campaign`: an `invite` step with a note template, a `wait` step with `untilAccepted`, then a `dm` and a `follow_up`. It runs on Finchreach’s servers, a reply from the lead stops the sequence for them, and the invite step can withdraw invitations that aren’t accepted after a number of days you choose.

Campaign templates use their own variables, with double braces: `{{firstName}}`, `{{lastName}}`, `{{fullName}}`, `{{company}}` and `{{headline}}`, each with an optional fallback such as `{{company|your team}}`. Any other variable is rejected, so details like the event or the post you read can’t be merged per person. That is a real trade-off: campaign notes are more generic than a hand-written one. Keep the template comfortably under 300 characters so long names still fit, and check `linkedin_preview_campaign` before launching.

> Ask your assistant: “Create a LinkedIn campaign for the people I add: invite with a short note that uses their first name and company (fallback “your team”), wait until accepted, withdraw after 21 days, then a thank-you message asking about their work, and one follow-up 7 days later. per_message approval. Show me the preview first.”

Sends from a campaign go through your account’s queue, inside the working hours and daily caps you set for the account (`linkedin_get_account_schedule` and `linkedin_update_account_schedule`). The full walkthrough is in [how to run a LinkedIn outreach campaign from your AI agent](https://finchreach.com/blog/linkedin-outreach-campaign-ai-agent).

## A note on safety

Automating a personal LinkedIn account carries risk under LinkedIn’s User Agreement, and no tool can promise your account won’t be restricted. What lowers the risk is what makes invitations good anyway: a person approving each one, a human pace, and notes written for one person at a time. More in [LinkedIn automation without getting banned](https://finchreach.com/blog/linkedin-automation-without-getting-banned).

## LinkedIn connection request messages: FAQ

What is the character limit for a LinkedIn connection note? Up to 300 characters (LinkedIn has given some free accounts a shorter limit, so check what your note box allows), and Finchreach enforces a 300-character limit on `linkedin_send_invitation` and campaign invite notes.

Should I always add a note? No. Add one when context helps: you haven’t met, or the reason isn’t obvious from your headline. Skip it when they expect your request or you’d only write filler.

How long should I wait before withdrawing an invitation? A few weeks is a reasonable default. Withdraw when you no longer expect an answer, not as a way to send the same invitation again.

Can my AI agent send invitations without me seeing them? Not one-off invitations: `linkedin_send_invitation` always waits in the approval queue. Campaigns let you choose: review every send, approve the templates once, or send without review.