How to research a prospect on LinkedIn before you reach out

Open any LinkedIn inbox and you can sort the messages into two piles in about three seconds each. One pile was written to a job title. The other was written to a person. The first pile gets archived without a reply. The second pile gets read, and sometimes answered, because the sender clearly spent a few minutes finding out who they were talking to before they typed a word.

Prospect research is the part of outreach that everyone agrees matters and almost nobody does consistently, because it feels slow and the payoff is invisible until the replies start landing. This post is about making it fast and repeatable: what to actually read on a LinkedIn profile, which signals predict a response and which ones are noise, and how to convert five minutes of reading into an opening line that could only have been written to that one person.

None of this requires a paid tool or a data vendor. It requires knowing where to look and what you're looking for. Once the pattern is in your head, a solid research pass on a single prospect takes three to five minutes, and most of that time is spent deciding whether the person is worth contacting at all.

Why research is the cheapest leverage in outreach

Personalization is overprescribed and underexplained. People hear it and reach for the lazy version: dropping the company name into a template, or referencing the prospect's city. That isn't research, it's a mail merge, and recipients can smell it instantly because the same sentence would work for ten thousand other people.

Real research changes the math in two ways. First, it tells you whether to send the message at all. A large share of prospects on any list are mistargeted, between people who left the company, took a different role, or were never a fit in the first place. Catching those before you message protects your reply rate and your reputation. Second, when the person is a fit, research gives you a specific, earned reason to be in their inbox, which is the only thing that reliably beats a delete.

The cost is low and front-loaded. You spend the minutes once, and the output feeds your first message, your follow-ups, and your read on whether this account is worth pursuing at all. Compare that to the cost of burning a connection request on someone who was never going to convert, and the research pass looks like the best trade in the funnel.

Start with the headline, not the job title

The first thing to read is the headline, the line directly under the name, not the formal job title from their current role. The two are often different, and the difference is informative. Job title is what HR calls the position. The headline is what the person chose to say about themselves to the entire platform, which means it's a small, deliberate piece of self-positioning.

Someone whose title is Head of Growth but whose headline reads 'Helping B2B SaaS teams fix their activation funnel' is telling you exactly what they care about and how they want to be seen. That phrase is a gift. It tells you the language they use, the problem they're closest to, and the angle most likely to make them stop scrolling. The title would have told you none of that.

Read the headline for three things: the specific problem they name, the audience they serve, and any claim or metric they chose to lead with. If the headline is just a job title and company, that's a signal too. It often means a less active user who treats LinkedIn as a directory listing, which changes how you should reach them and how long they might take to respond.

Read the 'About' and recent activity for intent, not biography

The About section and the activity feed are where you find the difference between someone who happens to fit your criteria and someone who is actively thinking about the problem you solve. You are not reading for biography. You are reading for intent and timing.

In the About section, look for what the person frames as their current mission rather than their career history. The last paragraph is usually the most current. In the activity feed, the goal is to find something they posted or commented on recently that connects to your offer, because a message that references a thought they shared last week proves you read it and gives them a natural reason to reply. Comments are underrated here: people are often more candid in a comment than in a polished post.

A word of caution on activity: recency matters. Referencing a post from two years ago reads as scraping, not reading. If their most recent activity is months old, lean on the profile and company signals instead, and don't pretend a stale post is fresh.

The signals that actually predict a reply

Not every detail on a profile is worth the same. After enough outreach you learn that a handful of signals do most of the predictive work, and the rest is texture. Here is the rough priority order, strongest first.

1. A recent role change. Someone who started a new job in the last three to six months is rebuilding their stack, their team, and their vendor relationships. They are unusually open to a relevant conversation, and the timing gives you a non-creepy reason to reach out.
2. A public company event. Funding, a product launch, a new office, or a hiring spree all signal motion, budget, and pressure. These are easy to verify on the company page and they make the 'why now' part of your message write itself.
3. Recent, relevant activity. A post or comment that touches your problem space, dated within the last few weeks, is the strongest individual-level personalization hook you can find.
4. Stated responsibility for your problem. If their headline or About names the exact outcome you affect, you don't need a clever angle, you need a clear one.
5. Mutual connections or shared groups. Weaker than the rest, but a genuine mutual connection lowers the cost of a connection request and gives a light, honest opener.

Notice what isn't on the list: their alma mater, their hometown, the fact that you both like a particular sports team. Those can warm a conversation that's already going, but as a cold opener they read as filler, because they have nothing to do with why you're worth their time. Lead with relevance, not rapport.

Qualify the company, not just the person

A profile tells you about a human. The company page tells you whether that human can act on what you're selling. Both matter, and skipping the company check is how good outreach lands on people who have no authority, no budget, or no problem.

On the company page, scan headcount and growth trajectory, recent posts for launches or funding, and the jobs tab. The jobs tab is one of the most honest intent signals on the platform: what a company hires for tells you where it's investing and what's currently broken. A company hiring three SDRs is making a different bet than one hiring its first RevOps lead, and your message should reflect which one you're looking at.

This is also where you catch the disqualifiers early. Wrong company size, an industry you don't serve, a region you can't support, an obvious in-house alternative. Disqualifying takes seconds and saves you from a polished message sent to someone who could never say yes, which is the most expensive kind of outreach there is.

Turn research into a message, not a dossier

The trap at the end of a good research pass is showing your work. You found six interesting things, so you put all six in the message, and now it reads like a background check. The recipient feels watched, not understood. Research is for you. The message should carry exactly one earned insight, stated plainly, with a clear reason you're reaching out now.

The structure that holds up: one specific observation that proves you read their profile, one sentence connecting that observation to a problem you can affect, and one low-friction ask. That's it. The research that didn't make the cut isn't wasted, it informs your follow-ups and your read on the account, but it stays out of the first touch.

If you want the message frameworks themselves, our guide to writing connection requests at warmerly.com/linkedin-outreach/connection-requests covers the opener formats that survive the first three seconds, and the LinkedIn message templates at warmerly.com/linkedin-outreach/templates show how to drop a single research insight into a structure without sounding like a script. Research and copy are two halves of the same job: the research finds the truth, the template keeps you from burying it.

Doing this at volume without losing the thread

Five minutes per prospect is fine for ten people. For a hundred a week it becomes the whole job, and the temptation is to either skip research or hand the entire outreach motion to a tool that blasts identical messages. Both paths end in the same place: a low reply rate and an account that starts getting flagged.

The realistic middle is tiering. Spend the full five minutes on your highest-fit accounts, the ones where a closed deal is worth real money, and run a lighter two-minute pass, headline plus company check, on the broader list. Batch the research separately from the writing so you're not switching modes every two minutes, which is what makes it feel slow. And keep your daily send volume inside what LinkedIn tolerates, because the best-researched message in the world does nothing if your account is restricted; our breakdown of LinkedIn connection and message limits at warmerly.com/linkedin-outreach/limits covers the current ceilings and the warning signs of moving too fast.

There's also the question of the account itself. A brand-new or long-dormant LinkedIn profile that suddenly sends dozens of connection requests looks like exactly what it is, and the platform treats it accordingly. Warming the account first, by establishing normal activity and ramping volume gradually, is what lets your researched outreach actually reach people instead of sitting in a restricted-account purgatory. Warmerly handles that warmup for LinkedIn alongside email, so the profile sending your carefully researched messages reads as an established, active human rather than a fresh automation. If you want to understand where automation helps and where it quietly hurts, the overview at warmerly.com/linkedin-outreach/automation draws that line.

A worked example

Suppose you sell onboarding software and you land on a profile: title Director of Customer Experience, headline 'Obsessed with the first 30 days. We win or lose customers in onboarding.' The About section's last line mentions she joined four months ago to rebuild the post-sale motion. Her most recent post, eleven days old, asks her network how they measure time-to-value. The company page shows two open roles for onboarding specialists.

You now have a stack of signals that all point the same direction: a recent hire, with explicit responsibility for your exact problem, actively asking the question your product answers, at a company visibly investing in the area. The first line writes itself, and it isn't 'I see you're the Director of CX.' It's a single sentence about her time-to-value question, connected to one concrete thing you've seen work, ending in a genuinely easy ask. Five minutes of reading, one message that could only have been sent to her.

That's the whole discipline. You're not looking for everything. You're looking for the one true thing that makes contacting this person now, with this message, obviously reasonable. When you can't find it, the research did its job too, by telling you to spend your next five minutes on someone else.