LinkedIn DM sequence writer: a free account gets a handful of personalised notes a month, so the sequence has to ration them

---
name: linkedin-dm-sequence-writer
description: Produces a complete LinkedIn outreach sequence for a named target list: which targets earn one of a free account's scarce personalised connection notes, which get a note-free invitation, which get followed and engaged with first, and the three messages that follow an acceptance. It writes to the correct character ceiling for the account type, checks each opener against a substitution test, and states the sending pattern that keeps the account inside documented invitation limits. This skill should be used when a target list exists but no outreach has been sent, when a connection note template needs writing or shortening, or when an account has been restricted and the sending pattern needs rebuilding.
---

# LinkedIn DM sequence writer

## The claim this skill is built on

The most repeated piece of LinkedIn outreach advice is "always personalise your connection request". On a free account that instruction is not executable, and almost nobody writing outreach advice has noticed.

As checked in August 2026, LinkedIn documents that free and Basic members can attach a personalised note to only a small handful of connection requests each month. It will not settle the figure: its most recently updated Help page says three per month, two older pages say five. Premium members are unlimited. Do not trust any source that states a single flat number here, this file included, without checking the product.

The character ceiling differs by account type too: 200 characters for free or Basic, 300 for Premium. The figure quoted almost everywhere is a flat 300, which is the Premium number, so a template written to it and pasted into a free account's note box loses its ending, which is usually where the ask sits.

Two consequences reshape the job. Personalisation is a rationed monthly resource rather than a default setting, and at roughly one note a week you cannot spend it evenly across a list of forty. And running out does not stop you: LinkedIn documents that restrictions and limits on standard invitations are separate from the limits placed on personalised invitations, two systems rather than one, so a note-free invitation remains available after the allowance is gone.

So what follows is not a template. It is an allocation rule, a taxonomy of what justifies an approach at all, and a message shape that survives rarely getting to introduce yourself in the invitation.

## Part one. Read the account's budget before writing a word

Every number below has to be read off the live product, because none of it can be inferred and one of them is unpublished entirely.

**Note allowance.** Free or Basic: a small handful per month, documented by LinkedIn as either three or five depending on which of its own pages you read, with the most recently updated one saying three. Plan to three and treat a fourth as a bonus. Premium: unlimited.

**Note length.** Free or Basic, 200 characters. Premium, 300. Write to 200 unless the account is confirmed Premium, because 200 works everywhere and 300 works in one place.

**Weekly invitation ceiling.** Commonly reported as around 100 invitations in a rolling seven-day window, and not documented. LinkedIn confirms limits exist and apply to Basic and Premium members alike but has never published a number, so treat roughly 100 as a widely observed working figure rather than a rule you can cite. Whether any paid tier raises it is disputed, and LinkedIn's own page says all members are subject to invitation limits.

**What crossing the line costs.** LinkedIn documents that an account sending too many invitations is temporarily restricted from sending them, typically for one week. While restricted you cannot send invitations, withdrawing pending invitations does not lift the restriction, you cannot buy your way out, LinkedIn cannot shorten the wait, and Support will not tell you the type or reason for it. Its own escalation guidance runs from waiting a few hours after a first restriction to waiting up to a month where the cause was too many outstanding invitations.

**Two more documented rules that change sequencing.** After withdrawing an invitation you cannot re-invite that person for up to three weeks, so withdrawing to free headroom costs you the target for most of a month. And a first-degree network is capped at 30,000 connections, at which point Follow becomes the default action on your profile and no invitations can be sent or accepted until you remove some.

**Acceptance rate.** LinkedIn lists invitations that were ignored, left pending, or marked as spam as a cause of restriction, which is the real basis for the advice that acceptance rate matters. No threshold is published and LinkedIn will not disclose the reason for a restriction, so any specific percentage you have read is somebody's marketing. Treat low acceptance as a documented risk with no published trigger point.

**Automation.** LinkedIn documents that if you send an excessive number of invitations and it suspects the use of an automation tool, it may suspend or restrict the account, and that repeated suspensions may result in permanent restriction. This is outreach, not automation. Everything below is written to be sent by a person, and the volume tooling that promises to remove the effort is what triggers the restrictions above.

## Part two. Ration the notes, and tier the list before you draft

Tier the whole list first. Drafting first creates sunk cost, and the first three names on the spreadsheet end up with the month's entire allowance for no reason other than being first.

Score each target on two axes, both answerable from the list itself. **Value:** does this person own the problem you solve, or can they only pass it along? Owning the budget or the remit is the bar, not seniority. **Trigger:** is there something specific and recent that makes an approach make sense today rather than at any point in the past two years? Part three defines this precisely.

Then place each target in one of five tiers.

**Tier 1, worth a note.** High value, a live trigger, and a specific evidence line that fits under 200 characters. Cap this tier at the month's allowance, so three on a free account. If more than three qualify, rank by trigger freshness and demote the rest rather than stretching the allowance.

**Tier 2, note-free invitation.** High value, but the trigger is weak or the allowance is spent. Send it bare. A bare invitation from a plausible profile is a normal event on LinkedIn; a note that says nothing is a worse first impression than no note at all, and on a free account it costs a quarter of the month's supply.

**Tier 3, follow and engage first.** High value, no trigger yet. Follow rather than invite, and wait for one. Engaging meanwhile only helps if the engagement is visible: LinkedIn documents that comments are ranked rather than chronological, that some appear only under the most recent view, and that it may limit the visibility of comments where it detects excessive commenting or an automation tool. A comment in the hidden tier warms nobody.

**Tier 4, InMail.** Premium only, where there is no realistic connection path and the value justifies a credit. See part six.

**Tier 5, do not approach.** No trigger, no ownership of the problem, or an obvious mismatch. Removing these is the cheapest thing you can do for acceptance rate, which is the documented risk to the account.

So a month on a free account is about three notes, some bare invitations kept well inside the observed weekly ceiling, and a growing follow list. If the plan needs more, it needs a different channel, not a faster sending pattern.

## Part three. What actually counts as a trigger

A trigger is a fact about the target that makes today a better day to write than yesterday. Most things treated as triggers are just facts about the target.

**Triggers that justify an approach.**

- They published something specific in roughly the last fourteen days that you have an informed reaction to, not a compliment about.
- They changed role into a remit that owns the problem you solve. Good for around ninety days; after that it is no longer news to them.
- Their company did something that changes the problem: entered a market, announced a migration, opened a location, published a job specification describing the work you do.
- They asked a question in public that you can answer without selling anything.
- A mutual contact has explicitly agreed to be named.
- You were both at the same event and you can name the session.

**Non-triggers that outreach tooling presents as triggers.**

- A work anniversary or a birthday.
- A profile view notification.
- Being in an industry, a city or a job title.
- Sharing mutual connections.
- Appearing in a filtered search export, which is a fact about your tooling.
- A milestone the platform surfaced to everyone who knows them, where your congratulation is the ninetieth.

The distinction is not sentiment. It is whether the fact is about them specifically or about a category they fall in. Triggers decay: a post trigger is stale after roughly two weeks, a role change after roughly three months. Approaching on a stale trigger reads worse than approaching cold, because it shows you noticed and then waited.

## Part four. The evidence line, and the test that makes it checkable

"Personalise the opening" is unfalsifiable advice. Replace it with a test.

**The substitution test.** Take the first sentence of the note or opening message and substitute another name from your list into it. If the sentence is still true, it is not personalisation, it is a greeting with a variable in it. Cut it and write one that breaks. "I really admire the work you are doing in fintech" survives substitution across half a list. "Your point about deprecating the legacy import path before the migration rather than after is the opposite of what most teams do" does not survive it at all, and that is the bar.

Three grades of evidence, best first:

1. A specific claim they made, quoted or closely paraphrased, plus your reaction to it: an extension, a counter-example, or a question showing you followed the argument.
2. A specific detail from their profile or their company's public material, plus the consequence you drew from it.
3. Nothing checkable, in which case the target is not tier 1. Demote it rather than dressing a category fact up as a personal one.

At 200 characters the evidence line and the reason for connecting share the space, which is roughly two short sentences. That is why tier 1 stays small: most targets have no fact about them that is both specific and compressible into one sentence.

## Part five. The three-message shape, and why message one has no ask

Once an invitation is accepted, the sequence is three messages. Not five, not a drip.

**Message one, the acknowledgement. No ask.** The invitation was the ask and it was just granted. A second request before any signal that the first was welcome is what makes people close threads. This message says why you are here, specifically, and stops. If you sent a note, extend the thought rather than repeating it; they already read it. If you did not, the evidence line goes here, and there is room: LinkedIn documents no character limit for a standard direct message at all, and the commonly reported 8,000 is observed rather than official. Ignore both numbers and write short, because the constraint is attention, not the platform. Attach nothing. A first message carrying a link, a calendar or a document is a pitch wearing an introduction.

**Message two, one small ask.** Sent once the first has had time to be seen, and only if it was read or replied to. The ask should be the smallest one that still moves the relationship: a specific question, a document worth opening, or a fifteen-minute conversation with a stated purpose. One ask, not a menu, because choosing is work and two asks reliably produce none.

**Message three, the close-out.** One message, and it ends the sequence. It says what you were going to offer, that you will not keep writing, and that they are welcome to reply later. No fourth message. This is not a trick to force a reply, it is what keeps you off the list of senders whose invitations get marked as spam, a documented cause of restriction.

**On spacing.** LinkedIn publishes nothing about message timing, so any interval is a choice rather than a mechanic. A defensible default is several days between messages, each conditional on the previous one being seen rather than firing on a clock. A clock-driven sequence sent to somebody who never opened message one is indistinguishable from automation.

**One documented safety net.** A sent direct message can be edited, text only, or deleted within 60 minutes, and an edited message carries an "Edited" label. Use it for a wrong name or a broken sentence. Two cautions: that window is a messaging rule and does not apply to comments, which are routinely conflated with it, and a sent InMail cannot be deleted at all, only edited before acceptance or within the same 60 minutes. An InMail mistake is permanent in a way a message mistake is not.

## Part six. The decision rule

For each target, in order:

- **They are already a first-degree connection.** Skip the invitation and go straight to a direct message carrying the evidence line. On a free account this is the only messaging path there is, since InMail is a Premium feature and free members can message only their connections.
- **High value, live trigger, an evidence line that breaks under substitution, and notes remaining this month.** Send a personalised note. 200 characters on a free account, 300 on Premium.
- **High value, live trigger, but the allowance is spent or the evidence line is weak.** Send a bare invitation. Note limits and standard invitation limits are separate systems, so this stays available, and the sequence picks up at message one on acceptance.
- **High value, no trigger.** Follow, do not invite. Engage where the engagement is substantive and revisit when a trigger appears. Most of a good list should sit here, and this is the branch skipped because it does not feel like activity.
- **High value, no realistic connection path, Premium with credits.** InMail, under the constraints below.
- **You cannot tell, because the account type, the remaining allowance, the invitations sent this week, or the target's trigger is unknown.** Send nothing and guess nothing. Open the product, read the three account numbers, then read the target's recent activity. Guessing the account type is what produces a 300-character note truncated at 200, and guessing the weekly count is how an account walks into a one-week restriction it cannot appeal, shorten, or get a reason for.
- **Low value or no trigger.** Do not approach. Ignored and pending invitations are a documented cause of restriction, so every low-conviction send draws against the same account standing the good sends depend on.

**InMail constraints, since this branch carries the most rules.** Premium only. Subject line 200 characters. The body is the one figure LinkedIn contradicts itself on: its most recently updated page says 2,000 characters and two Recruiter pages say 1,900, so write to 1,900 for one number that fits everywhere. Credits are allotted monthly by tier, documented as five for Premium Career, fifteen for Premium Business, fifty for Sales Navigator Core and thirty for Recruiter Lite, and they expire after 90 days. A credit comes back when the recipient responds within 90 days, whether they accept, decline or reply, but not while it sits pending, so an ignored InMail is a real cost and a declined one is not. You cannot send a second InMail to the same member until they respond to the first, and a member who has opted out of InMail in their message preferences cannot be reached this way at all.

## Part seven. Worked example, compressed

A founder sells a scheduling tool to multi-site veterinary practices. Free LinkedIn account, target list of eighteen exported from a saved search, no invitations sent this month.

**Budget.** Assume three notes, not five. Note ceiling 200 characters. No InMail available. The observed weekly ceiling is not binding at this list size.

**Tiering.** Four targets are practice group operations directors who own scheduling. Six are single-site practice managers who use a rota but buy no software. Five are veterinary surgeons with no operational remit. Three are suppliers. Among the operations directors: one posted eight days ago about losing locum cover to double-booking, one started in the role five weeks ago, one has posted nothing in two years, and one is already a first-degree connection.

**Allocation.** Tier 1 gets two, not three: the post trigger and the role change, both with evidence lines that break under substitution. The silent operations director goes to tier 3, follow and wait, because inventing a trigger would spend a third of the allowance on a category fact. The first-degree connection skips the invitation and gets a direct message. The six practice managers go to tier 2 as bare invitations spread across the month rather than sent in one afternoon. The surgeons and suppliers go to tier 5 and come off the list.

**Note for target one, 194 characters.** It names the double-booking point from their post, adds one concrete thing the founder has seen work at multi-site groups, and asks to connect. No product name, no link, no meeting request.

**Sequence after acceptance.** Message one extends the locum-cover thought and asks nothing. Message two, several days later and only once message one is read, offers one thing: a two-page breakdown of how three practice groups handled rota conflicts across sites. Message three, a week later, closes out in four lines.

**Verdict.** Two personalised notes, one direct message to an existing connection, six bare invitations spread over the month, one follow with no invitation, eight names removed. The allowance goes to the two targets where a note carries something no other note on the list could, and the remaining sixteen decisions are made without needing one.

## Failure modes

**The evenly spread allowance.** All three notes go to the first three rows of the spreadsheet, which are the first three alphabetically. From outside it looks like three generic notes followed by three weeks of bare invitations to better targets, exactly the wrong way round.

**The 300-character note on a free account.** The template was written to the figure published almost everywhere. On a free account it stops at 200, and since the ask sits at the end, what arrives is context with no request attached. Nothing reports this; the note just reads as unfinished.

**The category fact wearing a personal one.** "As a fellow operator in the veterinary space" survives substitution across the entire list. It is a greeting with a variable in it, and readers who get four a week recognise the shape instantly.

**The pitch as message one.** The first message after acceptance carries a link, a calendar and a product name. The acceptance is withdrawn in practice if not in fact: the thread closes, and often the connection goes with it.

**The clock-driven sequence.** Messages two and three fire on schedule at somebody who never opened message one. To the recipient that is indistinguishable from automation, which LinkedIn documents as grounds for restriction.

**The withdrawal spiral.** Facing a restriction, the sender withdraws pending invitations to bring the count down. LinkedIn documents that this does not lift the restriction and that a withdrawn invitation cannot be resent for up to three weeks. The account is still restricted and the list is now damaged.

**The volume tool.** Sending is handed to a browser extension or an outreach platform to remove the effort of rationing. It produces all the others at once, and LinkedIn documents that suspected automation can lead to suspension or restriction, with repeated suspensions capable of restricting an account permanently.

## What this skill does not do

- It cannot see your account, so it cannot tell you how many notes you have left this month, how many invitations you sent in the past seven days, or whether you are already restricted. The cannot-tell branch exists because guessing those three is the most common way this goes wrong.
- It does not research targets. With no access to their posts, profile or company news, the evidence line is only ever as specific as the material you supply.
- It does not send, schedule or automate anything, and should not be wired to something that does. The restrictions described throughout are triggered by exactly that.
- It cannot confirm the current limits. LinkedIn gives the free note allowance as both three and five on different live pages, gives the InMail body as both 2,000 and 1,900, and has never published the weekly invitation ceiling at all. Every figure here was checked in August 2026.
- It does not know whether an approach is welcome or legal where you are. Unsolicited commercial messaging is regulated differently across markets and none of that is assessed here.
- It cannot fix the reason nobody replies. If the offer is unclear or the profile does not support the claim in the note, no allocation rule rescues the sequence.