Warm leads in B2B sales,
and how long they last.

A warm lead is a person who took a recent, relevant action you can name and date, like a comment on a post about your problem.

Ask five sales teams what a warm lead is and you get five answers, most of which describe a spreadsheet rather than a person. A lead that matches the ideal customer profile. A lead that scored above 80. A lead from an account someone once met at a conference.

None of those describe warmth. They describe fit, and confusing the two is why so many "warm" sequences read exactly like cold ones.

This piece is definitional on purpose. What warmth actually is, how it differs from cold and hot, where the numbers everyone quotes really come from, and the two parts almost nobody plans for: how fast warmth wears off, and how much volume you surrender to chase it.

What actually makes a lead warm

A lead is warm when the person took a recent, relevant action of their own: a comment, a reply, a profile view, a registration. Warmth is a property of their behaviour inside a short window, not a field you computed. If you cannot name the action and the date it happened, the lead is cold.

The test is deliberately strict, because it is the only version that survives contact with a real inbox. Warm outreach works because the first line of your message can refer to something the recipient remembers doing. A lead score of 92 gives you nothing to open with. It ranks a list. It does not hand you a sentence.

So the two questions that decide whether a lead is warm are simple. What did they do, and when did they do it. If either answer is missing, you are writing a cold message with a warm label on it.

Cold, warm and hot, defined by who acted last

The three labels only mean something when you tie them to who moved last. Cold means nobody has acted, or only you have. Warm means they acted recently, in public, about your problem. Hot means they acted recently and aimed it at you. Everything else is packaging.

The useful consequence of defining it this way is that warmth stops being a mood and becomes a fact with a timestamp on it. You can audit it. Pick ten leads your team has tagged warm and ask for the action and the date on each. In most pipelines a good share of them turn out to be cold leads at good-looking companies.

Fit is not warmth, and a score is neither

Fit and warmth are different axes, and conflating them is the most common definitional error in B2B pipelines. A lead scoring 92 on firmographics is a good target, not a warm one, because nothing has happened yet. Scores rank a list you already had. Warmth tells you which row deserves a message this week.

Both axes matter, and they answer different questions. Fit answers "should we ever sell to this company", which is a stable property that barely changes month to month. Warmth answers "is this person paying attention to this problem right now", which changes by the day.

A high-fit, cold lead is a research target. A low-fit, warm lead is a distraction that feels productive. Only the high-fit, warm lead is worth a message today, and that intersection is much smaller than most target lists imply. Scoring the two axes separately, rather than averaging them into one number, is the whole argument of signal-based selling.

There is a second failure mode worth naming: warmth invented on your side of the relationship. Opening a tracked email, sitting in a nurture sequence, being uploaded to a retargeting audience, appearing in a purchased intent feed. Those are things done to the prospect, or measured about them, not chosen by them. They can be useful inputs. They are not warmth, because there is nothing you can reference in a first line.

The plain test: could you say the sentence out loud to the person without embarrassing yourself? "I saw your comment on that post" passes. "Our system noticed you opened three of my emails" does not.

How long a lead stays warm

Two clocks run at once, and most teams read the wrong one. The nurture clock measures how long a CRM record stays worth working, and the day counts printed against it in sales guides are convention rather than published measurement, so we are not going to repeat one. The reference clock measures how long the action stays worth mentioning out loud, and on LinkedIn it usually runs 24 to 72 hours.

Almost every page ranking for this question quotes the nurture clock and then gives advice that only makes sense on the reference clock. That is the mismatch that produces the four-week nurture flow triggered by a comment. The record is still worth working at week four. The comment is not still worth mentioning.

The windows below are the reference clock. They are working guidance rather than measured constants, and the useful part is the last column: warmth does not vanish, it downgrades into something else.

Two things fall out of that table. First, the windows are shorter than most sequence builders assume. A four-week flow triggered by a comment is not warm outreach, it is cold outreach with a delayed fuse, because by the time touch three lands the recipient has no memory of the thing you are referencing.

Second, the job change row is the odd one out, and the reason it runs long is instructive. The underlying condition, a person rebuilding their stack in a new seat, persists for weeks rather than seconds. Attention-based signals do not persist like that, which is the real rule behind the table: a signal stays referenceable for roughly as long as the state that produced it lasts. A comment is a moment. A new job is a season.

Job change signals are worth handling on their own cadence for exactly this reason, and the broader taxonomy of LinkedIn engagement signals sorts the rest by what each one actually tells you.

Where the warm‑lead numbers actually come from

Most warm-lead statistics are either unsourced or attached to the wrong study. The 5 to 15% conversion band in wide circulation has no primary source we could trace. The famous five-minute rule is real, but it belongs to a 2007 phone study of inbound form fills, not to the Harvard Business Review article it is usually credited to.

We went looking for the origin of each number you will see quoted. Here is what survived.

Three things are worth pulling out of that table.

The speed-to-lead literature is about inbound. Both Oldroyd studies measured what happens when somebody fills in a form on your website and you follow up. That is a person who raised their hand at you, which by the definitions above is a hot lead, not a warm one. The transferable finding is directional, that response value decays steeply and companies are far slower than they think. The specific multipliers do not transfer to a comment left under somebody else's post.

The two studies get fused constantly. The 21x qualification figure comes from the 2007 platform analysis. The 2,241-company audit with the 42-hour average comes from the 2011 Harvard Business Review article. They share an author, which is presumably how they merged in the retelling, but they are different samples measuring different things, and citing the audit numbers to the 2007 study or the multipliers to Harvard Business Review is wrong in both directions.

The audit is the underrated half. A 42-hour average first response across 2,241 companies, with 23% never responding at all, is a more useful fact than any multiplier, because it tells you the bar is on the floor. Same-day is not a stretch goal. It is the difference between you and roughly three quarters of the market.

Put the idea to work

Paste a post.
Get everyone who commented.

The comments finder reads a public post and lists the people under it, with what each one said. Free to try.

Find who commented

What warmth is actually worth, and the volume it costs you

Warmth multiplies the per-message result and divides the number of messages you can send. Trigger-based notes accept at 50 to 60% against 15 to 25% for generic ones, but you only get as many trigger-based notes as you found triggers. Run the arithmetic and cold at ceiling volume often still produces more conversations.

Here is that arithmetic, using LeadRiver's April 2026 acceptance figures and the practitioner-consensus weekly invitation ceiling. The one input nobody can hand you is how many genuinely warm people who also match your ideal customer profile a watch list turns up in a normal week. No published study measures it, it moves with your market and your list, and it is the number most teams never sit down and count. Count yours for two weeks before you plan around it.

Read the middle row carefully. The warm lane wins by roughly three times per attempt, and it only wins on absolute conversations once the signal supply gets large. A full cold lane produces 15 to 25 accepted conversations, so a warm-only lane has to turn up somewhere between 25 and 50 qualified signals every week just to match that, depending where each rate lands inside its band. Whether your watch list does that is an empirical question about your market that nobody has published an answer to, so measure it rather than assume it.

That is the tradeoff nobody states, and it produces a specific and very common failure: a team switches to warm-only prospecting, every dashboard metric improves, acceptance doubles, reply rate climbs, and the meeting count goes down. Nothing on the report is lying. The base just shrank faster than the rate grew.

The conclusion is not that warmth is overrated. It is that warmth is a targeting and copy instrument, not a volume strategy. The correct use is the bottom row: keep the volume lane running at whatever ceiling you have chosen, and let signals decide which people fill it and what the note says. You are not replacing cold outreach with warm outreach. You are warming the top of the cold lane.

Note also where the evidence stops. LeadRiver measured acceptance by note type, not what happens after acceptance. Belkins measured replies for existing connections against cold sequences, not replies to freshly connected signal-sourced people. Nobody has published the number you actually want, which is reply rate on a trigger-based first message to a new connection, so any model of it is an assumption and should be labelled as one. The campaign-level breakdown of what has been published lives in our cold versus warm outreach analysis, and the acceptance side is broken out further in our connection acceptance benchmarks.

For context on the channel itself, Expandi's state of LinkedIn outreach report for 2026, published in January 2026 across more than 70,000 campaigns, puts the LinkedIn platform average reply rate at 10.3%. The 5.1% cold email figure usually quoted beside it is Belkins research that Expandi cites, not Expandi's own data, and crediting it to Expandi is exactly the kind of laundering the table above exists to catch. That is a real gap, but a smaller one than the gap between a trigger-based note and a generic one on the same platform. The channel is not the variable doing the work. Belkins has since restated its cold email benchmark at 0.45% on 7.5 million 2025 sends, measured against total sends rather than opens, so treat 5.1% as the 2026-vintage figure Expandi quoted.

What to do when the warmth has expired

Do not send the warm message late. Referencing a comment from six weeks ago reads as surveillance rather than attention. Move the lead back to the list, keep the note about what they actually think, and wait for the next action to arrive, because for most people in most markets it does.

This is where the last column of the decay table earns its place. An expired signal has residual value, just not the value you originally wanted. You now know that this person holds a specific opinion about a specific part of your problem space, which is genuinely useful the next time they surface: your first line can be built on their position rather than on your product.

What does not work is treating a stale signal as if it were fresh. The two common versions are the sequence that fires four weeks after the trigger, and the manual message opening "I saw a while back that you...". Both draw attention to the delay, and the delay is the part that feels automated.

How to catch warmth before it expires

The constraint is not knowing what to say, it is noticing in time. Warmth on LinkedIn expires on a 24 to 72 hour clock, and nobody checks notifications reliably for a month. So the job is to watch a small number of posts and accounts continuously, then send on the same day.

Practically, that means three things.

Pick a small watch list. Ten to twenty accounts, plus the handful of people in your space whose posts your buyers actually comment under. A watch list you can hold beats a comprehensive one you abandon in week two, which is the usual outcome. It also sets your warm-lane ceiling, which is the number the table above depends on.

Read the engagement, not just the post. The people commenting under a relevant post are the warm ones, not the author. You can get that list without any account connected: our comment finder returns who replied under a given post, and the LinkedIn likes scraper does the same for the lighter signal.

Close the gap between noticing and writing. Most warmth is lost in the handoff, where the signal sits in a document until someone has an hour to write messages. If the same-day message is not going out, the watch list is decorative. Signal-based selling on LinkedIn covers the response frameworks that make same-day sending realistic rather than aspirational.

BeReach exists to compress that middle step. It finds who posted on your topic this month, used to work at a customer, or works at a company that just raised money or hired, lists a post's commenters, builds the list with each person's full details, and drafts the first line around the actual signal. A human reads and approves every one before it sends. The weekly number from the table above is still the one that should govern your plan.

Bottom line

A warm lead is a person who did something recent you can name and date, so act inside the window and use warmth to choose who fills your volume lane, not to replace it.

Frequently asked questions