LinkedIn Buying Signals: Which Ones Actually Predict a Reply in 2026
TL;DR: LinkedIn buying signals are public actions that suggest someone has a live problem. They are not equal. One person asking for a recommendation out loud is worth more than a hundred profile views, because the ask cost them something and the view cost them nothing. Rank every signal by that, act inside 48 hours, ignore the rest.
Most posts on this topic hand you fifteen signals and leave the ranking to you. The ranking is the whole job. Treat them as equal and you spend your morning on people who followed your company page while the one person who asked for a vendor in public gets answered by somebody else.
What LinkedIn itself counts as a buying signal
Start here, because LinkedIn publishes its own list and it is broader than most people assume. The Sales Navigator help page on activities that indicate buyer intent counts following a company, visiting a company page, viewing your profile, connecting with your colleagues, engaging with your ads, accepting an InMail, and visiting your website if you have the Insights tag installed.
Read that list again. Most of it is attention, not intent. A profile view tells you someone saw your name. A page follow tells you someone wants the content. Neither says a problem is live or that this person sits anywhere near a budget.
Not a knock on the product. Rolled into an account-level score, weak signals are fine for spotting a company warming up over a quarter. They are the wrong instrument for picking who to message on Tuesday morning.
Rank LinkedIn buying signals by cost to produce
The sorting question: how much did this action cost the person who took it, and do they sit near the money?
Cost beats volume as a proxy. A profile view is free and half accidental. Writing four sentences under a post about your category takes thought and a reason.
Tier 1: the public ask. Someone posts or comments asking for a recommendation, a tool, or help with the exact problem you solve. Strongest signal on the platform, and the only one where the buyer volunteered the timing. Also the most competitive, so the clock matters more here than anywhere else.
Tier 2: substantive engagement on category or competitor content. Not a like. A comment that names a constraint or compares two tools. That person is researching in public, and someone asking a competitor about switching costs is further along than anyone in your saved search.
Tier 3: trigger events with a budget attached. A new VP in the function you sell to, a job posting for the role your product supports, a funding round. Strong on timing, weak on problem. A window is opening; nobody necessarily has the pain yet.
Tier 4: repeat engagement from one account. Three people from the same company engaging over a month is a real pattern and a weak individual signal. Treat it as an account warming up, then go find the Tier 1 or Tier 2 person inside it.
The noise tier: profile views, page follows, drive-by likes. Worth a glance at who it was and nothing more. Messaging on a bare profile view is the LinkedIn equivalent of cold calling someone because they drove past your office.
The uncomfortable part: the strongest signals are the rarest. A handful of Tier 1 and Tier 2 signals in a week, not thirty. That is fine, because three good conversations beat forty ignored messages, and how to improve LinkedIn reply rates lays out the arithmetic.
Signals decay faster than anyone plans for
Every signal on that ladder has a short half-life. A comment is live context for about 48 hours, after which referencing it reads as surveillance rather than conversation. A job change holds for roughly 90 days, while the new person is still allowed to change things. A funding round holds for a quarter.
A list keeps. A signal does not. So four short mornings beat one long Friday, because three of them catch something inside its window.
The qualification pass to run before you send anything is in how to find buyers on LinkedIn.
The part I learned by doing this by hand
I built LeadBase because I was running this process by hand for my coaching business. Every morning I opened LinkedIn, went through the posts my buyers actually read, wrote down who had liked and commented, and pasted the names into my CRM one at a time. Forty five minutes, and the least sophisticated thing in my stack.
Two surprises. The messages wrote themselves, because when you reply to something a person said out loud twelve hours ago there is no personalization step to fake. And the habit was first off the calendar in a heavy client week, so my pipeline quality tracked how busy I was rather than how good my method was.
One thing took longer to see. Sorting by engagement volume is the wrong sort: the post with the most likes reliably gave me a worse list than the post with the most comments. Same cost-to-produce rule, learned the slow way.
Log them or you are guessing
The failure mode with LinkedIn buying signals is not missing them. It is never finding out which kind paid. Three fields on every conversation you start:
- What the signal was (public ask, comment on category content, job change)
- Where you saw it (whose post, which alert, which company)
- What happened (no reply, reply, call booked)
Run it for thirty days, then count booked calls by signal type rather than by message template. Almost nobody does, which is why advice in this category stays generic.
There is vendor data pointing the same way, with the caveat that it is a vendor counting its own platform rather than an independent study. Expandi’s State of LinkedIn Outreach report, drawn from more than 70,130 campaigns run through its tool, reports a 13.4% reply rate on campaigns that message people who already viewed your profile, against 8.62% for template-driven campaigns.
Now notice what that does and does not say. A profile view is the weakest thing on my ladder, and it still beat a blind template blast. The floor is not whether the signal is strong, it is whether the person did anything at all. That is an argument for working the top of the ladder first. It is not an argument for building your week around profile views.
Where tooling helps, and where it does not
Tooling handles the boring half: watching more posts than you can read, catching engagement inside its window, keeping the log. That category is broken down in LinkedIn engagement tracking tools.
It does not handle the judgment half, though the market sells it as if it does. No tool knows whether a comment described a real constraint or was written to farm impressions, and none of them should write the reply. The advantage of a signal is that you read the thing, and the other person can tell.
One more trap: enrichment is not intent. An email address makes somebody reachable, not interested, and LinkedIn data enrichment for B2B covers why the two get sold as one product.
FAQ
What are LinkedIn buying signals? Public actions suggesting a person or company has a live problem you could solve. They run from an explicit request for a vendor recommendation down to a profile view.
Which LinkedIn buying signals are strongest? A public ask for a recommendation, then a substantive comment on content about your category or a competitor. Both cost real effort, which is what separates them from a like or a follow.
How fast do I need to act on a buying signal? About 48 hours for engagement signals, roughly 90 days for a job change, a quarter for a funding round. Engagement goes stale fastest, because the person stopped thinking about the post.
Do I need Sales Navigator to find LinkedIn buying signals? No. Its buyer intent score is an account-level tool for spotting companies warming up over time. Finding the individual worth messaging today means reading comment sections, which is free.
What to do next
Want to see how this works before changing anything? Take a look at LeadBase and how it surfaces engagement signals from the posts your buyers read.
Prefer the manual version first? How to find warm leads walks the by-hand process I ran for months before I built anything.
Ready to stop rebuilding the list every morning? Start a free LeadBase trial and let it watch the signals while you do the judgment part.
