What warm outbound on LinkedIn actually requires
Warm outbound on LinkedIn requires a specific, nameable prior connection that the buyer would also agree exists. That last clause is the whole test, and most outbound described as warm fails it.
The practical version of the test is uncomfortable and quick. Write down, in one sentence, why this buyer would not consider your message cold. If the sentence names a person, a conversation, something they published or a group you both genuinely belong to, the message is warm. If it describes research you did, a trigger event you spotted in a tool, or the fact that they fit your ideal customer profile, it is cold, and the buyer will experience it as cold no matter how it is phrased.
This matters because the two produce different correct behaviour. Cold outbound at volume is a legitimate motion with its own arithmetic, and it works when the market is large and the deal size supports the hit rate. Warm outbound is a low-volume motion that trades reach for reply quality. Problems start when a team runs the first and describes it as the second, because the framing then has to be manufactured, and a manufactured pretext is worse than no pretext at all.
GTM Brigade is a LinkedIn engagement platform for B2B GTM teams that builds curated buyer watchlists, drafts comments in the founder's voice, routes buying signals to reps, and attributes the resulting pipeline in HubSpot and Salesforce. This guide covers the outbound half of that motion, written so a team could run it without any tooling.
The four sources of warmth, in order of strength
Four sources produce real warmth: a mutual person who will vouch, a direct prior interaction, a public signal the buyer created, and shared membership of something specific. They are not interchangeable, and each earns a different opening.
A vouch. A mutual contact who will actively say something on your behalf, or allow you to say they suggested you get in touch. This is the strongest source because it transfers trust the buyer already holds rather than asking them to extend new trust to a stranger. It is also the scarcest, and the reason most warm outbound runs on weaker sources. The condition is that the mutual contact genuinely agreed; borrowing a name without asking converts the strongest source into the most damaging one.
A direct prior interaction. You and the buyer have exchanged something: they replied to your comment, you spoke at an event, they answered a question you asked. This is strong and verifiable, and its strength decays with time in a way the first source does not.
A public signal the buyer created. They published a post, announced a role change, shared a problem, spoke on a podcast. You are responding to something they chose to make public, which is a legitimate opening even though you have never interacted. This is the most abundant source and the most commonly overstated: responding to a post is warm, having noticed a funding round is not.
Shared membership of something specific. The same community, cohort, alumni group or customer base. Specific is doing the work: both being in B2B software is not membership, both having been in the same twenty-person cohort is.
| Warmth source | Strength | Abundance | What it earns |
|---|---|---|---|
| A vouch from a mutual contact | Highest | Scarce | A direct ask, named referrer |
| Direct prior interaction | High | Limited | A continuation of the exchange |
| Public signal they created | Medium | Abundant | A response to that signal only |
| Specific shared membership | Lower | Moderate | An introduction, context first |
| Profile research | None | Unlimited | Nothing. This is cold outbound |
Why a manufactured pretext costs more than it buys
A pretext the buyer recognises as manufactured does more damage than an honest cold message. They now know two things: that you wanted something, and that you were willing to dress it up.
This is the most expensive mistake in the motion and it is made constantly, usually in good faith. A rep has no genuine warmth source, knows a cold message will probably be ignored, and so constructs one: a compliment about a post they skimmed, a claim to have been following the buyer's work, a reference to a shared interest lifted from the profile. Each individual instance is harmless-looking. The cumulative effect is that experienced buyers now read every warm opening as a probable pretext, which is why the genuinely warm ones have to work harder than they used to.
The honest alternative performs better than teams expect. A message that says plainly that you have not met, states specifically why you are writing to this person rather than to anyone, and asks one small question, is received as straightforward. Buyers are not offended by being sold to. They are offended by being handled.
Forrester's research on B2B buyers has described a growing preference for self-directed evaluation and a corresponding impatience with seller contact that adds nothing, which is the mechanism underneath this. The objection is not to outreach. It is to outreach that costs the buyer attention and returns no information.
There is a version of this that is subtler and equally costly: a real source, overstated. You did comment on their post once, three months ago, and the message implies an ongoing relationship. The source exists, so the rep feels honest, and the buyer still experiences the mismatch between the claim and their own memory. The rule that handles both cases is to state the source at exactly its real strength and let the message be as modest as that requires.
So the operating rule is narrow: name the source, or send the message without a source and say so. Never invent one, and never promote one.
The three instruments, and what each one permits
A connection request with a note, a message to an existing connection and an InMail are three different instruments with three different acceptable openings. Using the opening appropriate to a stronger instrument than you hold is the second most common error here.
The connection request with a note. One or two sentences, stating the warmth source and nothing else. No pitch, no link, no question that requires work. The note's only job is to make accepting the request an easy decision. A blank request is weaker and honest; a request carrying a pitch spends the single moment of reliable attention you get on something unrequested.
A message to an existing connection. The widest latitude, because they already accepted you. This is where a specific question or a short ask belongs. The trap here is the opposite one: an existing connection is not an existing relationship, and treating a two-year-old accepted request as a warm source produces the pretext problem above.
InMail. A paid instrument that bypasses the connection step, and the buyer knows it was paid for. That raises the relevance bar rather than lowering it. It is the right choice when the source is strong and the timing is genuinely urgent, and the wrong choice as a way of getting around a weak source, which is mostly how it gets used.
One practical note on volumes: warm outbound is low volume by its nature, so teams that run into platform restrictions are almost always running volume outbound with warm framing. If the process needs automation to hit its numbers, the framing is doing no work except in your own reporting. The signal and sequencing side of this is set out in the LinkedIn engagement-to-pipeline playbook for B2B GTM teams.
What the first message should actually do
Ask for something smaller than a meeting. The purpose of the first message is to convert recognition into an exchange, and a calendar request treats warmth as currency to be spent rather than as a relationship to continue.
The structure that works has three parts and fits in four sentences. Name the source in the first sentence, specifically enough that the buyer can place it. State why you are writing to them rather than to anyone in their position, which is the sentence that proves you are not running a list. Then ask one question whose answer takes under a minute and that you would genuinely find useful.
What that achieves is a reply, and a reply changes the relationship category. You are no longer someone who contacted them, you are someone they have corresponded with, and the meeting ask from inside a conversation is a different and much easier request. Gartner's research on the B2B buying journey describes buyers moving through independent investigation in their own order, which means the useful thing to be is available and credible when they arrive at the relevant stage, rather than insistent before they do.
Two things to leave out of the first message. Any attachment or link, because both ask for more commitment than a reply and both look like a sequence. And any reference to the buyer's company performance, growth or funding, because it signals that they were selected by a filter, which undoes the work the warmth source just did.
How to tell whether it is working
Watch reply rate and the share of replies that reference your stated source. The second number tells you whether the warmth was real, which is the thing you actually want to know.
Reply rate alone is ambiguous at low volume, and warm outbound is always low volume, so a single week tells you almost nothing. Over a quarter it becomes meaningful, and the useful comparison is against your own cold outbound rather than against a published benchmark, because the benchmark was measured on a different list with a different motion.
The diagnostic number is the second one. When buyers reply and mention the mutual contact, the comment, the post or the cohort, your sources are genuine and legible. When replies come back polite and generic, the sources are probably thinner than your messages claim, which is the pretext problem showing up as data rather than as a complaint.
One failure pattern is worth watching for because it looks like success. A team switches to warm outbound, reply rate rises sharply, and everyone concludes the motion works. Then nothing converts. What usually happened is that the messages got politer rather than better sourced, and polite messages earn polite replies from people who have no intention of buying. The guard against it is to check whether replies progress to a second exchange, not merely whether they arrive, because a one-reply conversation is a courtesy rather than a signal.
A third thing worth tracking is where the warmth came from, recorded per message. After a quarter you will know which of the four sources actually produces conversations for your market, and that answer differs enough between businesses that no guide can give it to you. For how any of this connects to recorded pipeline in a CRM, including why the decisive interaction is invisible to click attribution, see the guide to attributing LinkedIn to pipeline.