LinkedIn Prospecting: The Full Process, Step by Step
LinkedIn prospecting is the process of finding qualified buyers on LinkedIn, contacting them directly, and managing those conversations until they become meetings. It works in five stages: define the list, respect the weekly invitation limits, open with a message tied to a real signal, track every r
Most articles on this topic are lists of tips. Tips are useless here, because LinkedIn prospecting fails in a specific order: people get the list wrong, then blame the message. This is the process in the order that the failures actually happen.
Why does LinkedIn prospecting stop working for most teams?
It usually fails for one of three reasons: the list is too broad, so no message can be relevant to all of it; the sending volume trips LinkedIn’s weekly invitation cap and the account gets restricted; or replies arrive and nobody owns them, so warm conversations go cold inside an inbox.
Notice that only one of those three is a copywriting problem. Teams spend most of their effort rewriting the first message, which is the cheapest thing to fix and rarely the thing that is broken. A perfect message sent to the wrong 2,000 people still fails.
The order below is deliberate. Do not move to step two until step one is done, because every later step multiplies whatever quality the list has.
Step 1: How do you build a LinkedIn prospect list worth messaging?
Start from a signal, not a job title. A workable list is defined by something that just happened — a funding round, a new hire in a relevant role, a technology added, a job posting that reveals a gap — plus the title of the person who owns that problem. Lists built on title and company size alone produce reply rates in the low single digits.
A practical size check: if you cannot write a first line that is true for every person on the list, the list is too broad. That is the whole test. It takes thirty seconds and it saves weeks.
In practice this means smaller lists than most people expect. A list of 200 people who share a real signal outperforms a list of 3,000 filtered by industry, because the second one forces you into a message so generic that it reads as a template — and LinkedIn users recognise a template instantly.
Two filters worth applying before you send anything. First, remove anyone who has not been active on LinkedIn in the last month: an inactive profile is a wasted invitation against your weekly cap. Second, remove anyone already in your CRM as an open opportunity, because a cold approach to a live deal is the fastest way to annoy an account executive.
Step 2: How many connection requests can you send per week?
LinkedIn enforces a weekly cap on connection requests, and the exact ceiling varies by account age, acceptance rate and whether you hold a premium subscription. The practical rule is to stay comfortably below whatever your account tolerates and to watch your acceptance rate: a falling acceptance rate is what triggers restrictions, not volume alone.
This is the step where most accounts get damaged, and the damage is slow. A restricted account does not announce itself; invitations simply stop being accepted and nobody notices for a fortnight. Check the cap and the current rules before you set a daily number.
«LinkedIn’s weekly connection request limit»
One thing that matters more than the raw ceiling: withdraw pending invitations that are older than two or three weeks. They count against you while they sit there, and clearing them is the single easiest way to improve acceptance rate without changing anything else.
Step 3: What should the first message actually say?
The first message should name the signal you found, state in one sentence why that makes you relevant, and ask for something smaller than a meeting. Three to four lines. No attachment, no calendar link, no pitch. The goal of message one is a reply, not a booking — optimising for bookings at this stage is what produces the 1% reply rates people complain about.
The most common mistake is asking for fifteen minutes in the first line. It is a bigger ask than it looks: you are asking a stranger to give you time before they know whether you understand their problem. Ask a question they can answer in one sentence instead.
The second most common mistake is personalisation that is obviously automated. Mentioning that someone went to a particular university, when it has nothing to do with why you are writing, signals a tool, not attention.
«LinkedIn prospecting message examples that get replies»
Step 4: How do you keep track of replies without losing them?
LinkedIn’s inbox is not a pipeline. It has no stages, no owner field, no follow-up dates and no way to see how many conversations are waiting on you. Once more than one person is prospecting, or once you pass roughly fifty active conversations, replies need to sit in a CRM that records stage, owner and next step.
This is the stage nobody plans for, and it is the one that wastes the most money. The cost of losing a warm reply is not the reply — it is the whole list, the sending time and the weeks of waiting that produced it. Teams routinely generate good conversations and then lose a third of them to an unread inbox.
The symptom is easy to spot. If anyone on the team has ever said “I think I already messaged this person,” the tracking layer is missing.
«a CRM built around LinkedIn conversations»
Step 5: Which metrics tell you whether it is working?
Track four numbers weekly: acceptance rate on invitations, reply rate on first messages, positive reply rate, and meetings booked. Connections added is not a metric — it measures activity, not progress. If reply rate is healthy but meetings are not, the problem is the second message, not the list.
The diagnostic value comes from reading them together, not separately:
Low acceptance rate
The list or the profile is the problem. People are judging you before they read anything, which means your headline, photo and recent activity are doing the work — badly.
Fix the profile before touching the message.
Good acceptance, low reply rate
The first message is the problem. They accepted you, so they were open; they just did not find a reason to answer.
This is the only case where rewriting the message is the right move.
Good reply rate, no meetings
The second and third message are the problem. You opened a conversation and then could not move it.
Usually a case of pitching too early on message two.
Everything good, no pipeline
The list is qualified to talk but not qualified to buy. Right person, wrong moment or no budget.
Go back to step one and change the signal, not the volume.
What are the most common LinkedIn prospecting mistakes?
The four that cost the most: sending to a list too broad to write a relevant first line for; asking for a meeting in message one; treating the LinkedIn inbox as a pipeline; and measuring connections instead of replies. All four are process mistakes, not writing mistakes.
«the LinkedIn prospecting mistakes we see most often»
Do you need Sales Navigator to prospect on LinkedIn?
No, not to start. Sales Navigator earns its cost when you need saved searches with change alerts, lead lists shared across a team, or filters that the free search does not expose. For a single person working a list of a few hundred people built from a clear signal, the free search plus a decent outreach tool covers the same ground.
«Whether Sales Navigator is worth it»
What tools does LinkedIn prospecting need?
Three layers, and most teams over-buy the first and skip the third:
Finding and qualifying
LinkedIn search or Sales Navigator, plus whatever gives you the signal you chose in step one — funding databases, job boards, technology detection.
Cheapest layer to get right and the one that decides everything downstream.
Sending and sequencing
A tool that sends invitations and follow-ups on a schedule, respects the weekly limits, and stops a sequence automatically when someone replies.
That last feature is not optional. A sequence that keeps sending after a reply is worse than no sequence.
Tracking and handing off
Where replies become pipeline: stage, owner, next step, and a record that survives the person who started the conversation leaving the company.
The layer teams discover they needed about four months in.
If you are comparing tools in this category, the honest way to do it is by which layer each one actually covers, because most cover one and a half and market as if they covered three.
«the GTM tool stack we recommend»
What we learned running this on ourselves
One Pexio campaign, run to 28 September 2026: 766 invitations to Sales Directors at companies of 50 to 200 employees, 421 accepted, 147 replied and 20 meetings booked. Moving the company-size filter up from 1 to 50 employees raised replies by about 20% with no change to the messages.
We run Pexio on Pexio. Here is one campaign, from 1 of September to 28 September 2026, with the numbers and the mistake in the middle of it.
The target was Sales Directors at companies of 50 to 200 employees. The sequence was a connection request followed by two messages. Over the month the list reached 766 people.
The funnel, in absolute numbers: 766 invitations sent, 421 accepted, 147 of those people replied, and 20 meetings booked. That is a 55% acceptance rate and a 35% reply rate — one meeting for every seven people who answered, and one for every 38 people on the original list.
The part worth your time is what happened halfway through.
We started with companies of 1 to 50 employees. Partway in we moved the filter to 50 to 200, and replies rose by about 20%. Nothing else changed — same target role, same three messages, same person sending them.
The reason turned out to have nothing to do with size. Companies in the larger band have an actual sales team. A Sales Director at a twenty-person company frequently is the sales team, and a tool that changes how a team coordinates outreach solves a problem they do not have yet. Above fifty employees there are people to coordinate.
That is a targeting lesson, not a copywriting one, and it is the reason this article puts the list before the message. We rewrote nothing. We changed who received it.
One caveat we would rather state than have you notice: we switched the segment mid-campaign instead of running both at once, so some of that 20% could be timing or sample size rather than the segment itself. We are running it as a clean split next quarter and will update this number.
«a full LinkedIn prospecting case study»
Frequently asked questions
How long does it take to see results from LinkedIn prospecting?
The first replies arrive within days, but a reliable read on whether the approach works takes four to six weeks: enough time for two or three sequence cycles and for the acceptance and reply rates to stabilise. Judging a list in week one is judging noise.
Is LinkedIn prospecting better than cold email?
They answer different problems. LinkedIn gets higher visibility and lower volume; email gets higher volume and easier automation. Teams that run both usually use LinkedIn to open the relationship and email to follow up, because reply data ends up in the same pipeline either way.
How many messages per day is safe on LinkedIn?
There is no universal number, because the ceiling depends on account age, premium status and above all acceptance rate. The safe practice is to start low, watch the acceptance rate weekly, and withdraw invitations that have been pending more than two or three weeks.
Can LinkedIn prospecting be automated?
The sending and sequencing can be. The list and the first line cannot be, in the sense that automating them is exactly what produces the generic messages that fail. Automate the schedule and the follow-ups; keep the signal and the opening human.
What reply rate should I expect from LinkedIn prospecting?
Our experiencie shows us that it depends on industry, weekly or monthly posts added to the profile and product or service offered. Our results, based on the tool we sell, Pexio, is 55% on average for invitationes accepted and 30% of messages answered. Usually we try to get 20 meetings per month per user.
Want to apply this in your company?
Pexio automates your LinkedIn prospecting so you can focus only on closing sales.