LinkedIn InMail vs Connection Message: Which to Use
Default to the connection request — it is free and turns your next message into a first-degree one. Use paid InMail only for people you cannot reach otherwise.
Default to the connection request — it is free and turns your next message into a first-degree one. Use paid InMail only for people you cannot reach otherwise.
That is the whole decision, and for most B2B outreach it is not close. An accepted invite changes your standing permanently: you become a first-degree contact who can message as often as the relationship justifies, at no cost. An InMail buys you exactly one delivered message and nothing more unless the person replies. The exception is real and worth paying for — but it is narrow, and it is defined by reachability, not by ambition.
Related: InMail vs cold email: picking the channel
The decision, by scenario
Find your row. The verdict column is the answer — there is nothing to weigh.
| Your situation | Use | Why |
|---|---|---|
| They are already a first-degree connection | Regular message | You already have a free channel; spending a credit on a contact buys you nothing |
| Standard B2B prospect you are able to invite | Connection request | Free, and an accept converts every later message into a first-degree message |
| Out-of-network, or LinkedIn will not let you invite them | InMail | It is the sanctioned way to reach someone the invite door is closed to |
| Their profile carries an Open Profile badge | Free message, no credit | LinkedIn lets anyone message Open Profile members without spending a credit |
| Senior executive who visibly does not accept cold invites | InMail | One precise paid message beats an invite that will sit pending indefinitely |
| Your well-targeted invite has gone unanswered for weeks | InMail as the fallback | The free path already had its turn; escalate once, deliberately |
| Big deal, short named list of accounts | InMail | Credits are scarce and a handful of named targets is exactly what they are for |
| High volume — hundreds of prospects a month | Connection requests | Credit allowances are small; only the free path scales to that shape |
| No paid LinkedIn subscription | Connection request | InMail requires a paid tier, with Open Profile members as the only exception |
| You want an ongoing relationship, not one reply | Connection request | An accept is permanent; an InMail is a single delivered message |
| Genuinely urgent — days, not weeks | InMail | Delivery does not depend on someone accepting you first |
| Warm context: a mutual, an event, a post they wrote | Connection request | The context already does the work you would otherwise pay a credit for |
Two patterns fall out of that table. Everything to do with reachability points at InMail. Everything to do with volume, budget, and durability points at the connection request. Nothing about the size of your ambition belongs in the decision.
Why the connection request is the default
It changes your standing, not just your delivery. This is the mechanism that decides the whole argument. An InMail is delivered as a paid message from a stranger and stays that way. An accepted invite makes you a first-degree contact — and from that moment, messaging is free, unmetered, and framed as contact-to-contact. You are not buying one delivery; you are buying the channel itself.
It is the only path that scales. InMail credits are allocated monthly and are small relative to any real outbound motion. Invites are limited by patience and by LinkedIn’s weekly ceiling, not by your subscription. If your list has hundreds of names on it, the connection path is not merely preferable — it is the only one that arithmetic allows.
The ask is smaller. “Do I want this person in my network?” is a lower-friction question than “do I want to reply to a stranger who paid to reach me?” That is a framing difference rather than a measured one, but it is the reason accept-then-message is the standard play at mid-market and below.
It compounds. Every accept is permanent. A year of invites leaves you with a network you can message for free forever; a year of InMail leaves you with a receipt.
The lever inside this path is the note. Attach one, and make it specific to them — one concrete observation about their work, their company, or something they published. The note field is short (commonly around 300 characters on a standard account as of 2026, with some paid tiers allowing more — check yours before writing to a fixed length), so a single relevant sentence about them outperforms a paragraph about you. Personalization is not only a reply lever; it is a safety lever, because relevant invites get accepted and irrelevant ones get ignored or reported.
When InMail wins
Reach for InMail when the free path is not available or has already failed. Four situations qualify:
1. You cannot send the invite at all. Distant-degree members, restricted invitation settings, and profiles LinkedIn simply will not let you invite. InMail is the only route LinkedIn gives you to those inboxes, and there is no free equivalent.
2. The invite will never be accepted. Some senior people run a closed network as policy. An invite to them is not a slow yes — it is a permanently pending invite that occupies one of your slots and returns nothing. Paying to be read once is strictly better than waiting forever for free.
3. The invite already had its turn. A well-targeted request that has sat unanswered for weeks has given you your answer about that channel. If the account is genuinely worth it, a short, respectful InMail is a clean escalation — one deliberate second knock, not a nag.
4. The clock is the constraint. InMail delivery does not depend on an acceptance step. When the window is days rather than weeks, that dependency is the risk you are paying to remove.
What InMail is not is a shortcut around targeting or writing. Because it arrives as a paid message from someone the recipient does not know, the bar for a reply is higher, not lower. Generic InMail reads like an advertisement — and unlike an ignored invite, it also costs you a credit to be ignored.
What each one actually costs
Money. Connection requests are free and need no subscription. InMail requires a paid LinkedIn tier — Premium Career, Premium Business, Sales Navigator, or Recruiter — and is not available on a free account, with Open Profile members as the standing exception. As of 2026, Sales Navigator Core is published at around $99.99/month billed monthly and is cheaper billed annually, with Advanced tiers costing more. Monthly credit allowances are small and tier-dependent: single digits on the cheapest Premium tier, a few dozen on Sales Navigator. LinkedIn changes tiers and allowances, so confirm both the price and your allowance on LinkedIn’s own pricing page before you plan around them.
Two published credit mechanics are worth knowing because they change how you should spend. Unused credits roll over, but the accumulation is capped — commonly around three months’ worth — so hoarding them has a ceiling. And a credit is refunded when the recipient replies within LinkedIn’s published window, which means a well-targeted InMail that lands is close to free, while a badly targeted one is not. That single rule is the strongest argument against volume InMail there is.
The cost that is not money. A connection request is free, but it is not costless.
| What you spend | Connection request | InMail |
|---|---|---|
| Cash | Nothing | A paid LinkedIn tier, plus one credit per send |
| Allowance | One of your weekly invite slots | One monthly credit, refunded if they reply in LinkedIn’s published window |
| Lingering commitment | A pending slot, held until they accept or you withdraw it | None — it lands or it does not |
| Account exposure | Real: acceptance rate and “I don’t know this person” responses are watched | Lower on invite signals, though generic paid messages can still be reported |
| What you get back | A permanent first-degree channel | One delivered message |
The play: sequence them, don’t choose between them
The strongest teams do not pick one channel. They order them, so the free option always goes first and the paid one is held in reserve for the cases it is genuinely good at.
Step 1 — send a personalized invite. One specific, relevant line. No pitch in the note; the goal of the invite is the accept, not the meeting.
Step 2 — wait properly. People accept on their own schedule, and a pending invite is a normal state rather than a failure. Impatience here is what pushes teams into behavior that gets accounts restricted.
Step 3 — message once they accept. Open with something human that references why you connected. Firing a pitch the second the acceptance notification arrives wastes the standing you just earned.
Step 4 — hold InMail in reserve. Spend credits on the two cohorts that deserve them: people who never accepted and are genuinely worth a second knock, and people you could never invite in the first place.
Run that as one sequence rather than four manual habits. In WarmySender, a LinkedIn campaign is a multi-step sequence with conditional branching — you branch on whether the invite was accepted and on whether they replied, so an accept routes the prospect onto the follow-up message and a reply stops the sequence immediately. Wait-for-accept and wait-for-reply steps hold the line between stages instead of a timer guessing for you. And because Send InMail is a first-class step alongside invites and messages, the fallback branch is part of the same campaign, not a separate manual chore. We recommend the connection path as the default for the mechanical reason above — not because the paid one is missing from the product.
If your prospects live in two places, put the whole thing in a multichannel sequence that mixes email and LinkedIn steps in one flow. The two channels cover each other’s blind spots: email reaches people who never accept invites, LinkedIn reaches people whose inbox filters you. Cold emailing, email warmup, LinkedIn outreach, and multichannel all run from the same workspace, so the sequence is one object rather than two tools stitched together.
The safety frame underneath all of it is what makes automating this survivable. Daily caps are enforced per action type — invites, messages, InMails, and profile views each get their own budget, so a busy invite day cannot starve your follow-ups. New accounts run on a progressive ramp that starts volume low and climbs over weeks. When an account runs out of InMail credits, InMail steps are deferred to the next monthly reset rather than retried against a wall, while invites and messages on the same account keep running. An InMail aimed at a non-Open-Profile member from an account that cannot send it is refused rather than attempted. And an account-health view shows usage against every cap, with warnings before you get near one. Defaults sit deliberately below the documented platform ceilings, because account safety wins over throughput every time.
Frequently asked questions
What is InMail?
InMail is LinkedIn’s paid direct-messaging feature that lets you message a member you are not connected to, without needing them to accept a connection request first. It requires a paid LinkedIn subscription and consumes one InMail credit per message, with credits allocated monthly by tier. Open Profile members are the exception: anyone can message them without spending a credit.
What is the difference between InMail and a regular LinkedIn message?
A regular LinkedIn message is free and only available between first-degree connections, while InMail is paid and reaches people you are not connected to. Once someone accepts your connection request, you can message them as often as the relationship justifies at no cost — which is why the accept is worth more than any single paid message.
Is InMail better than a connection request?
No, not as a default — the connection request is the better opener for most B2B outreach because it is free, uncapped by credits, and an accept converts every later message into a first-degree message. InMail is better in one specific circumstance: when you cannot reach the person any other way, whether because you cannot invite them, they will not accept cold invites, or your invite has already gone unanswered.
How much does InMail cost?
InMail has no per-message price tag; it requires a paid LinkedIn subscription and spends one credit per send, with a small monthly credit allowance that depends on your tier. As of 2026, Sales Navigator Core is published at around $99.99/month billed monthly and cheaper billed annually, with cheaper Premium tiers offering only single-digit credit allowances — check LinkedIn’s pricing page, since tiers and allowances change. Unused credits roll over up to a cap, and a credit is refunded if the recipient replies within LinkedIn’s published window.
Can you send InMail on a free LinkedIn account?
Not in general — InMail requires a paid tier such as Premium Career, Premium Business, Sales Navigator, or Recruiter. The one exception is Open Profile members, who can be messaged by anyone on LinkedIn, including free accounts, without spending a credit. If you are on a free account, the connection request is your channel.
Should I put a note on a connection request?
Yes, when you have something specific to say — one concrete, relevant line about their work, their company, or something they published. The note field is short (commonly around 300 characters on a standard account as of 2026, with some paid tiers allowing more), so a single sentence about them beats a paragraph about you, and a note that reads as genuine research protects your acceptance rate as well as your reply rate.
Can I send an InMail to someone I am already connected to?
You should not — messaging a first-degree connection is free and unlimited, so spending a credit on someone already in your network buys you nothing. Reserve credits for people outside your network entirely. If your outreach tool is sending InMail to existing connections, that is a targeting problem, not a channel problem.
The bottom line
Use the connection request. It is free, it scales, and an accept upgrades your standing permanently rather than buying you one delivery. Reserve InMail for the narrow set of people you cannot invite, will not be accepted by, or have already invited without an answer — and spend those credits deliberately, because a well-targeted one that gets a reply is close to free while a badly targeted one is not. Best of all, do not pick: sequence the invite first, message on the accept, and keep InMail in the fallback branch where it belongs.