LinkedIn Connection Request Templates That Actually Get Accepted (For SMB Outreach)

You are sending LinkedIn connection requests to your Dream 100 list and watching the accept rate sit somewhere between flat and embarrassing. The list is right. The platform is right. The bottleneck is the request itself.

This post is the fix. Three template structures that consistently get accepted, written to be copied straight into your outreach tool with two or three personalization tokens swapped in. Each one is under the 300-character LinkedIn connection-note limit. Each comes with a working version (the A copy) and a tighter version (the B copy) so you can see what to keep and what to cut.

For what happens after the prospect accepts, the four-message LinkedIn sequence is the companion read. For the full Pillar 4 view, the LinkedIn outreach hub gathers every spoke.

The Rule: Connection Request Notes Never Pitch

Before any template, the rule.

Owners and BDRs lose at the connection-request stage because they treat 300 characters like a mini sales letter. They lead with what they do, who they help, and a soft ask. The prospect reads two lines, smells "selling," and either ignores or declines.

The connection request is not the pitch. It is the doorbell. If the doorbell sounds like a salesman, the door does not open.

The job of the request note is to earn the accept. That is the whole job. Three sentences to prove you are a real person, that you looked at the prospect's profile or feed, and that you have a reason to be in their network beyond "selling to them." Nothing more. The pitch lives later, in message 1 of the four-message sequence after the connection is in.

If you find yourself wanting to add "I help companies like yours with..." into the note, stop. That belongs in message 1 (and even there, it is rarely the strongest move). Out of the note. Use this rule first; the templates only work inside it.

Template 1: The Mutual-Connection Opener

Best fit: you share at least one second-degree connection with the prospect, and the mutual is a real connection (someone you have actually worked with, talked to, or transacted with), not a LinkedIn ghost.

Why it works: humans accept requests faster when the request comes inside an existing trust graph. Naming the mutual signals you are inside the prospect's professional world, not parachuting in from a list buy.

Template 1A (the working version):

Hi [First Name], noticed we both know [Mutual Connection Name]. [Mutual Name] and I worked together on [one-sentence specifics: a project, a topic, a context]. Happy to connect over here. No pitch coming.

Template 1B (the tighter version):

Hi [First Name], saw [Mutual Connection Name] in our shared network. [Mutual Name] and I [one-line context, project or topic]. Would be good to be in your network. No pitch, just a connect.

What changed between A and B: B drops the "happy to connect" pleasantry and tightens the mutual-context line to a single clause. The structural beats are identical; B reads as more direct and runs about 14 characters lighter.

The "no pitch coming" close. Most owners cut this line because it feels redundant. Keep it. It does the same work as a hand wave in person, signaling the next message will not be a sales push. Prospects feel the difference.

Tokens that matter: [First Name], [Mutual Connection Name], the one-sentence mutual context. The mutual-context sentence has to be real and specific. "We worked on the spring rollout for a client in your space" beats "we have collaborated." If you cannot write the specific line, you do not actually know the mutual; pick a different template.

Use this when at least one strong mutual exists. If your mutuals are open networkers with 30,000 connections each, the trust signal is weaker; switch to Template 2 or 3.

Template 2: The Recent-Post Reference

Best fit: the prospect has posted on LinkedIn in the last 60 days and the post said something specific enough to react to.

Why it works: referencing a recent post proves you opened the prospect's profile and read something they wrote. That alone separates the request from the inbound that did not. The prospect's brain registers "this person actually paid attention" before the question of whether to accept comes up.

Template 2A (the working version):

Hi [First Name], your post on [specific post topic] caught my attention. The line about [specific phrase or idea from the post] matched what we are seeing across [their industry or role]. Connecting over here, no agenda.

Template 2B (the tighter version):

Hi [First Name], the post on [specific post topic] landed for me, especially the bit on [specific phrase or idea]. We see the same pattern in [their industry or role]. Connecting over, no agenda.

What changed between A and B: B replaces "caught my attention" with the more concrete "landed for me" and condenses the industry callback. The "no agenda" close stays in both.

The thing to avoid: "Great post!" by itself. The compliment without the specific quote-back is the signature of a generic outreach tool, and the prospect can smell it from the preview. The whole point of this template is the specific phrase from the specific post. If you cannot name the phrase, do not run this template.

Tokens that matter: [First Name], [specific post topic], [specific phrase or idea from the post], [their industry or role]. The specific phrase is the lever. Spend 20 seconds on the post; pull the actual line.

Use this when the prospect posts regularly and their feed has at least one real line worth referencing in the last 60 days. If the prospect's feed is dead, switch to Template 1 or 3.

Template 3: The Specific-Question Opener

Best fit: no strong mutual exists and the prospect does not post often, but their role or company has an obvious operational angle you can ask about. This is the workhorse template for the back half of any Dream 100 list.

Why it works: a specific question is the lowest-friction way to signal you are interested in the prospect's world, not selling into it. The question is not asked to be answered in the note (no one answers a question in a connection request). It sets the tone for the conversation that follows in message 1.

Template 3A (the working version):

Hi [First Name], came across your profile while researching how [their industry or role] teams are handling [specific operational topic] this year. Curious how it is playing out at [Company]. Connecting over here, no sales pitch attached.

Template 3B (the tighter version):

Hi [First Name], researching how [their industry or role] teams are handling [specific operational topic] right now. Wondering how it looks at [Company]. Connecting over, no pitch.

What changed between A and B: B drops the "came across your profile" opener (most prospects assume you found them via search; saying so adds nothing) and tightens the operational frame. The "no pitch" close is non-negotiable in both.

The trap to avoid: asking a question whose answer is your product. "Curious how you are handling marketing follow-up this year" sent by a marketing service is transparent. The question has to be one a peer might ask at a conference, not one a vendor would ask a buyer. If it reads like a discovery-call opener, rewrite it.

Tokens that matter: [First Name], [their industry or role], [specific operational topic], [Company]. The operational topic carries the weight. "How CFO teams in mid-market manufacturing are sequencing AP automation in 2026" is specific. "How you are handling growth" is not.

Use this when you can name a real operational topic with one sentence of specificity. If you cannot, the question itself is the bottleneck; go back to the Research role and find one before you send.

Personalization Tokens That Matter (And Tokens That Do Not)

The fastest way to make a connection request look automated is to personalize on tokens that mean nothing.

Tokens that matter (use these):

• First name. Always.
• Mutual connection name (Template 1 only).
• Specific post topic and a specific phrase from the post (Template 2 only).
• Specific operational topic (Template 3 only).
• Company name (Templates 1, 2, and 3, when natural).
• Role when the role is genuinely specific (CFO, Head of Operations, COO) and not a generic seniority label.

Tokens that do not matter (skip these):

• Generic city. "Saw you are in Dallas too" is a tell. Cities show up on every profile and signal that you scraped a field.
• Generic industry. "Fellow logistics person" reads as a mail merge.
• Generic compliment. "Impressive background" or "love what you are doing" is the universal automation signature.
• Job title verbatim with no context. "Saw you are VP of Sales" is a field pull, not a personalization.
• Years at company or anniversary congrats unless you are sending in the actual anniversary window and have something genuine to add.

Personalization is not the act of inserting a token. It is the act of proving you read something. A token swap without a read attached is the giveaway.

The two-second test: read your draft note out loud. If you can substitute any other prospect's name and company into the same sentence and it still reads as written, the personalization is fake. Rewrite. Use this test before every send.

The Accept-Rate Question (And How to Frame Yours)

The first question owners ask after reading this far is "what accept rate should I expect."

The honest answer: it varies more than any benchmark will tell you, and most published numbers come from sales-tool vendors with skin in the game. We are not going to cite a specific figure here.

The frames that hold up in practice:

• Cold connection requests with no personalization sit at the low end of any range you will see. Many lists drop into the single digits.
• Personalized requests using one of the three templates above consistently run several multiples higher than the same list run with no personalization. The gap is large and visible inside two weeks.
• Mutual-connection requests (Template 1) outperform the other two when the mutual is genuinely strong.
• Dead lists (wrong titles, wrong stage of company, wrong industry) do not get rescued by a good template. The list is upstream of the note. If your accept rate is flat with personalized notes, the list is the issue; go back to the Dream 100 list build post and look there first.

The takeaway: track your own accept rate per template for 30 days. That number is the only benchmark that matters. Published averages are interesting; your own number is actionable. Use this.

What To Do This Week

If you have a Dream 100 list and you are sending connection requests:

1. Save the six templates above (1A, 1B, 2A, 2B, 3A, 3B) into your outreach tool or a doc you can copy from.
2. Tag each prospect on your list with the template best suited to them: mutual exists (Template 1), active poster (Template 2), or operational angle (Template 3).
3. Send 25 requests using the matched templates this week. Run the two-second test on every one before you send.
4. Track accept rate by template across all 25.
5. Calendar a 15-minute review of your numbers at day 14 and again at day 30.

If you do not have a Dream 100 list yet, build that first. The templates only work on the right people, and the list is upstream of every note in this post.

If your accepted prospects go quiet, the four-message LinkedIn sequence is the next read. The connection request is the doorbell. The sequence is what happens after the door opens.

The Bigger Picture

LinkedIn outreach is one role inside a working marketing function. Research builds the list. Outreach sends the note. Follow-Up handles the replies. Content keeps your feed warm between touches. Reporting tells you which template is earning the accepts.

The owners who win on LinkedIn outbound are not the ones with the magic note; they are the ones with a team behind them keeping the lists current, the notes personalized, the messages sent on time, and the replies handled inside 24 hours.

That team is what your digital marketing team is built to be. Done-For-You on the cadence side, with you in the strategic seat.

When you are ready to stop running this yourself, we are here.

Precision Marketing Execution, Automated.