Why Your LinkedIn Messages Get Filtered: Connection-DM vs InMail Deliverability
By Elena Marsh, Strategy & Algorithm. Last updated: 2026-05-30
- You rewrite the message copy ten times and the reply rate barely moves.
- You buy InMail credits and watch them vanish into silence.
- You count "messages sent" as the metric, but never measure where they actually landed.
- Your reply rate keeps sinking below your send count and you cannot explain the gap.
What is the difference between a connection DM and an InMail for deliverability?
A connection DM rides an accepted relationship and lands in the recipient's primary inbox, while an InMail is a paid push to a stranger that LinkedIn (and the recipient) can route to the filtered Other tab. The two are not interchangeable.
When someone accepts your connection request, they have granted you standing. Your message arrives as a normal conversation in the focused inbox, ranked alongside messages from colleagues and peers. An InMail carries no relationship signal. It is a cold, paid interruption, and both the platform and the recipient treat it that way. Recipients who do not recognize the sender often let InMail sit in the Other tab unread, and a single dismissal teaches LinkedIn to deprioritize the next one. The practical takeaway is that the channel decides placement before your words get a vote. A great message in the Other inbox loses to an average message in the primary inbox. For a deeper read on when each channel is worth using, see Linked Insider: InMail vs connection request in 2026.
Why does so much of your send volume never get a reply?
Most of your send volume never gets a reply because the funnel leaks at three points, not one: acceptance, reply-of-accepted, and the silent drop where messages never surface. The reply gap is partly a placement gap.
Reframe the numbers. Across 316,703 LinkedIn outreach sequences run on the verified API, Reachium's data shows a 28% average connection acceptance rate. Of those accepted connections, 29% replied, which works out to roughly 8.1% of all connection requests sent. That last figure is the one demand-gen marketers should pin to the wall. You are not optimizing a reply rate against your send count. You are optimizing it against a much smaller pool that survived acceptance and placement first. When a marketer sees a low reply number and immediately rewrites the opener, they are often fixing the wrong layer. The leak is upstream, where sends fail to reach the primary inbox at all. The flagship breakdown of these benchmarks lives at Linked Insider: LinkedIn outreach benchmarks 2026. If your acceptance itself is the weak point, Linked Insider: why no one accepts my connection requests covers that layer directly.
Want to put this into practice?
Reachium automates LinkedIn outreach, content publishing, and inbox management in one platform.
Start Free →Connection DM vs InMail: which one actually lands?
A connection DM lands in the primary inbox and out-reply InMail in most B2B contexts, while InMail earns its credits only when you have no path to a connection and the recipient is genuinely worth a paid touch. Here is the head-to-head.
| Factor | Connection DM | InMail |
|---|---|---|
| Default inbox placement | Primary / focused inbox | Often the filtered Other tab |
| Relationship signal | Accepted connection (warm) | None (paid cold push) |
| Cost | Free within send limits | Consumes paid InMail credits |
| Volume ceiling | Calibrated to ~25 invites/day | Capped by your credit balance |
| Spam-report exposure | Lower (recipient opted in) | Higher (unsolicited stranger touch) |
| Reply benchmark | 29% of accepted reply (~8.1% of sent) | Typically lower, varies by seniority |
| Link tolerance in first touch | Low; links throttle delivery | Low; links throttle delivery |
| Best use | Repeatable, measurable outbound at scale | One-off reach to an unconnectable, high-value target |
The pattern is clear. InMail is a deliverability tax you pay for the privilege of skipping the connection step, and that skip is exactly what costs you primary-inbox placement. Use it sparingly, for targets you cannot reach any other way. For the response-rate side of that decision, Linked Insider: LinkedIn InMail response rate has the comparison, and Linked Insider: connection request message examples covers the warm path.
How do spam reports and link dampening suppress your account?
Spam reports and links trigger a suppression loop: reports lower your future placement, links and attachments in a first touch get throttled, and over-sending compounds both effects. The account quietly gets dampened before you notice.
LinkedIn's Professional Community Policies make clear that unsolicited, repetitive, or commercial messaging draws enforcement, and report signals feed the platform's filtering. Every time a recipient marks a message as spam or simply ignores a cold InMail, the system learns to route your next message lower. Links and attachments in a first touch make it worse, because a first message that asks the recipient to click before any trust exists reads as exactly the pattern filters are tuned to catch. Then there is the volume tax. Reachium's data shows acceptance peaked at 34% for accounts sending 10-19 invites a day and fell to 30.6% at 20-29 a day. More volume produced fewer accepts, which is why the platform caps sending around 25 a day by design. Pushing harder does not buy you reach. It buys you suppression. The publicly reported HeyReach ban in March 2026 is the cautionary version of this for browser-automation tools, where the failure mode is not a dampened account but a removed one.
How do you instrument delivered versus replied on LinkedIn?
You instrument delivered versus replied by tracking the full chain (send, to placement, to reply) instead of judging copy from reply count alone. Watch the Other-inbox ratio and read leading indicators before you touch the message.
Start by separating the layers you can act on. Track acceptance rate, then reply-of-accepted, then the share of conversations that surface in the primary inbox versus the Other tab. If your acceptance is healthy but replies are thin, the leak is placement or copy, and you can A/B those independently: run a version of the sequence with a link in touch one and a version with the link held to touch two, and compare. Timing is another leading indicator most marketers ignore. A message that lands when the recipient is offline gets buried under newer threads, so send-window data matters; Linked Insider: LinkedIn message send-time reply heatmap maps when replies actually cluster. The discipline is to stop guessing from a single aggregate number. When you can see where the funnel leaks, you stop rewriting openers to fix a problem that lives in placement.
Want to put this into practice?
Reachium automates LinkedIn outreach, content publishing, and inbox management in one platform.
Start Free →What send path keeps messages in the primary inbox?
A verified-API send path to accepted connections keeps messages in the primary inbox, because it routes through LinkedIn's sanctioned channel and respects the limits that trigger suppression. That protects both deliverability and the sending account.
The send path is the variable most marketers never consider. Tools built on browser automation or scraping mimic a human clicking, which violates LinkedIn's User Agreement and exposes the account to enforcement. A path built on the official API through a sanctioned partner like Unipile sends through the front door, so messages route the way LinkedIn's own product intends, into the focused inbox of an accepted connection. Pair that with pacing that stays under the volume-tax threshold and the suppression loop never starts. For a demand-gen marketer who needs delivered-versus-replied to be a real, measurable metric, the send path is the difference between optimizing copy and optimizing placement.
FAQ
What is the difference between a connection DM and an InMail for deliverability?
A connection DM is sent to someone who accepted your request, so it lands in the primary inbox as a normal conversation. An InMail is a paid message to a stranger that LinkedIn and the recipient can route to the filtered Other tab, which is why it usually delivers worse.
Why is my InMail being filtered as spam?
InMail carries no relationship signal, so unrecognized senders, commercial language, and first-touch links all push it toward the Other tab. Report and ignore signals from past recipients also lower the placement of your future sends.
Do links in a first message hurt delivery?
Yes. A link or attachment in a first touch, before any trust exists, matches the pattern LinkedIn's filters are tuned to throttle. Hold the link until a second message and compare your reply rates to confirm the lift.
How do spam reports affect future LinkedIn message delivery?
Each report or dismissal teaches LinkedIn to deprioritize your next message, and the effect compounds across recipients. Combined with over-sending, this suppression loop can dampen an account well before it produces a visible block.
How do I tell if messages are landing in the Other inbox?
Track the share of conversations that surface in the primary inbox versus the Other tab, and watch reply timing. If acceptance is healthy but replies are thin and delayed, placement is the likely leak rather than your copy.