In short
- 1"Defining, finding and qualifying cost nothing to redo. Only the send draws on a limited weekly allowance of invitations."
- 2"The scarce resource is invitations. Industry consensus puts the ceiling near 100 a week, and LinkedIn itself publishes no official number."
- 3"Across 48 workspaces and 42,236 connections, the mean response rate was 15.7% and the median 4.5% (BeReach, 2026)."
- 4"Notes anchored to something the person recently posted were accepted at 50 to 60 percent against 15 to 25 percent for generic ones (LeadRiver, 50,000+ requests, April 2026)."
Read five guides on this query and you get the same list five times. Optimise your profile, learn boolean search, buy Sales Navigator, personalise the note, follow up, go multichannel, measure. None of it is wrong. All of it is tactics without a budget.
What is missing from every one of them is the two things that actually decide whether the loop works. The first is that the four stages of prospecting do not cost the same, and three of them cost nothing at all. The second is an honest number for what comes back, because the figures those guides quote are averages, and the average team on LinkedIn is not the median team.
That is what this page is for. The tactics are covered in depth elsewhere on this site and linked from each stage below. Here we price the loop.
What LinkedIn prospecting is
LinkedIn prospecting is the loop of deciding who you sell to, finding those people on LinkedIn, checking each one is a fit, and reaching out. Three of those four stages are free and effectively unlimited. Only the reaching out is rationed, by LinkedIn's own weekly ceiling. That asymmetry should decide how you build the whole thing.
Most guides treat the four as a linear funnel where each step costs roughly the same effort. They do not. Stages one to three are reading and thinking, which you can do all day. Stage four spends a finite, non-renewable allowance that resets slowly, so a name spent on the wrong person is one you do not get back this week.
The loop, and what each stage costs
The whole method fits in one table. Stages one, two and three cost nothing to repeat, so run them exhaustively. Stage four is metered by the platform, so it should only ever receive people who survived the first three. Everything else on this page is elaboration on that.
The practical consequence is an ordering rule: every check that can disqualify someone should happen before the one that spends an invitation. Done in that order, being wrong about a prospect is free. Done in the usual order, being wrong costs an invitation you do not get back this week.
Stage 1: define who you sell to
Write down the segment before you open LinkedIn. A usable definition names the company shape you serve, the role that owns the problem, and the trigger that makes now the right moment. Without the trigger you have a list. With it you have a reason to write, which is the part that actually moves acceptance.
Two of those three are easy and everybody does them. Company shape and role are just filters, and filters are what search boxes are for. The trigger is the one people skip, and it is the one the data cares about. LeadRiver's April 2026 analysis of more than 50,000 connection requests found notes anchored to a specific piece of content the recipient had recently posted accepted at 50 to 60 percent, against 15 to 25 percent for generic or blank requests.
Note what that comparison rules out. First-name personalisation is not the lever. LeadRiver describes it as table stakes that produces no meaningful uplift over generic on its own. The uplift comes from having a reason, and a reason is a property of your segment definition, not of your copywriting.
If you want the taxonomy of which triggers are worth building a segment around, signal-based selling on LinkedIn covers the six that fire reliably, and intent signals you can track without Sales Navigator covers where to read each one.
Stage 2: find them
Finding people draws on public data: public profiles, public company pages, public job posts and public comments, and for most B2B offers that is enough to build the list. Degree filters, reactor identity and other Sales Navigator facets are the main things public data alone does not give you.
This is the point where the standard guide sends you to buy something. It is worth knowing what the paid lane actually adds before you do. Sales Navigator is genuinely good at headcount bands, seniority facets and saved-lead alerts. It is not what gives you a name, a title and an employer, because those are on the public profile.
Three finding routes carry different amounts of information, and they are not interchangeable:
- Search by role and company. The cheapest route and the least informative. It tells you who someone is, not whether they care. Start here to size the segment, not to pick who to write to. The free people search and free company search both read the public lane.
- Hiring activity. A company posting three roles in the function you serve has told you where its money is going this quarter. The free jobs search reads public job posts.
- Public comments under relevant posts. The richest route, because a comment is an opinion. Someone who wrote three sentences about your problem under somebody else's post has pre-qualified themselves in a way no filter can match. The LinkedIn comment scraper turns those comments into a list you can act on.
There is a subtlety worth getting right: commenters are public in a way reactors are not. The comment text sits on the page and tells you what the person thinks, while a reaction carries no such signal and a reliable identity behind it is harder to pin down. If you are choosing one to build a prospecting pass around, the commenters are both cheaper and better. Prospecting without Sales Navigator walks the full public-lane route.
Stage 3: qualify before you spend an invitation
Qualify before you spend an invitation, not after. Reading a public profile costs nothing and can be repeated all day. An invitation is one of roughly a hundred you get in a week, and it does not come back. Run every free check first, so that every rejection is free and only survivors cost you anything.
The usual order is backwards. Most teams find a name, send the invite, wait, and only start asking whether this person was a fit once the conversation goes nowhere. By then the invitation is spent and the answer arrives too late to be useful.
The checks worth running, cheapest first:
- Role and employer. On the public profile. Disqualifies most of a raw list in seconds.
- Company shape. Size band, sector, headquarters, all on the public company page.
- Trigger present or absent. A recent role change, an open job post, a comment on your topic. No trigger means no reason to write, which means the note will read generic no matter how it is worded.
- Already in your pipeline. This one needs your own records, not LinkedIn.
Only after all four does anyone earn an invitation. This is also where an AI agent earns its keep, because grading a few hundred public profiles against a written segment is exactly the kind of judgement that is tedious for a person and cheap for a model. BeReach runs that grading on public data, and needs only a substantive headline plus a resolved employer to produce a grade.
You can define a segment, build a list, grade every name against it and read the drafts before you commit to any of it. Deciding what actually goes out is the last step, not the first.
Stage 4: reach out
The send is the only stage with a ceiling. LinkedIn's consensus limit is about a hundred invitations a week. BeReach adds its own hard ceilings of 50 invitations a day and 120 profile visits an hour, with a three-minute minimum interval between two invitations and between two messages.
Two things about that consensus figure are worth stating plainly, because most articles get them wrong. LinkedIn publishes no official weekly invitation number anywhere. The roughly 100 a week figure is what practitioners converge on from observation, and any page that cites the LinkedIn Help Center for it is citing something that does not exist. Second, the platform limit and your tool's pacing are two separate layers, and the tighter one wins. The full breakdown of connection request limits covers how they interact and what happens when you push past either.
For completeness on the BeReach side: 300 profile visits a day and 70 messages a day are base figures that a workspace multiplier scales, so treat them as the floor of what a paced account will do rather than a promise about the maximum. The 50 invitations a day and the 120 visits an hour are unconditional and are not scaled by anything.
That budget makes the sequencing decision for you. A hundred invitations a week against a segment you have not qualified is a hundred coin flips. The same hundred against a shortlist where every name carries a trigger is the only version of this that compounds.
On the message itself, the shortest useful rule is that the note should be one the recipient could not have received from anyone else. Everything else is covered in what to write in a connection request, cold message templates that survive contact, and the broader outreach template library. Every message BeReach drafts is reviewed and approved by a human before it sends, which is the practical answer to the failure mode where volume and quality come apart.
What to expect, honestly
Published LinkedIn benchmarks all quote averages, and averages describe this population badly. Across 48 workspaces and 42,236 connections on our own platform, the mean response rate was 15.7% and the median was 4.5%. Plan against the median. The tenth percentile team got 0.4% and the ninetieth got 49.5%, so the mean describes almost nobody.
The last two rows are the reason this section exists. When the mean is more than three times the median, a small number of outliers are carrying the average and it stops being a planning number. The tenth to ninetieth percentile spread inside a single product, on a single platform, over the same period, was more than a hundredfold.
One more figure that no benchmark nets out: in that same sample, 68.3% of all responses were explicit disinterest. A reply is not a conversation. If you are modelling pipeline from a published reply rate, divide by three before you believe it. The full study, method and distribution has the percentile table.
This is not an argument against LinkedIn prospecting. It is an argument for building the loop so that the metered stage receives only qualified people, because at median performance you cannot afford to spend invitations on anyone else.
What LinkedIn prospecting actually costs
A working stack is usually a data layer plus a sending layer. Sales Navigator Core is $119.99 per month billed monthly, per seat. BeReach is priced per workspace rather than per seat: Pro is EUR 99 per month billed monthly, or EUR 948 a year, which works out at about EUR 79 a month, and includes two connected accounts.
Sales Navigator figures are from LinkedIn's compare-plans page, checked August 2026. Both tiers include 50 InMail credits per month per seat.
The per-seat versus per-workspace distinction is the one that changes the arithmetic on a team. A three-person sales team on Sales Navigator Core billed monthly is $359.97 a month before any sending tool. BeReach's price does not move with headcount, it moves with how many LinkedIn accounts you connect. There is a free tier with 200 monthly credits and a three-day trial if you want to run the first three stages before deciding whether the fourth is worth paying for. The AI is one included model with no picker and no key to bring, so there is no second bill hiding behind the first.
Every viral post is 100+ warm conversations waiting.
Tell your agent who you want to reach. It finds them, says which ones are worth your time, writes the first line, and follows up.
The mistakes that break the loop
The loop breaks in four predictable places: volume set before fit, qualification done after the invite instead of before, personalisation that is only a first name, and no measurement past reply count. Each one is cheap to fix and expensive to leave, because each one wastes the single resource you cannot buy more of.
Volume before fit. Deciding you will send 100 a week and then finding 100 people is the wrong order. Decide who qualifies, then send however many that produces. If the answer is 30, send 30.
Qualifying after the send. Covered above, and it is the single most common structural error. It converts a free decision into a paid one for no benefit.
First-name personalisation. LeadRiver describes this as table stakes that produces no meaningful uplift over generic on its own, and publishes no separate acceptance figure for it. The lift it does measure is 50 to 60 percent for notes carrying an actual reason against 15 to 25 percent for generic. The gap between "Hi Sarah" and a specific reference is where the whole result lives.
Measuring replies instead of interested replies. With 68.3% of responses in our sample being explicit disinterest, a reply-rate dashboard flatters a campaign that is failing. Count the conversations that continued.
Two topics sit outside the loop and are worth reading separately: LinkedIn's own rules on what triggers a restriction are covered in why LinkedIn restricts accounts and how to automate outreach within LinkedIn's limits, and the platform's own rules on how much searching a free account may do are in the commercial use limit.
Frequently asked questions
What is LinkedIn prospecting?
LinkedIn prospecting is the process of defining a target segment, finding those people on LinkedIn, qualifying each one against the segment, and reaching out. It differs from general lead generation in that the sending stage is metered by the platform, so the qualification stage has to happen first. Three of the four stages are free and unlimited. Only the send is rationed.
How many connection requests can I send per week?
Industry consensus puts the ceiling around 100 invitations a week, which is roughly 15 to 20 on a working day. LinkedIn publishes no official figure, so any page citing the Help Center for it is citing something that does not exist. BeReach caps a connected account at 50 invitations a day as a hard ceiling, with a minimum of three minutes between two invitations.
What is a good LinkedIn reply rate?
Published benchmarks put it at 7.9% for cold connect-then-message and 12.2% once connected (Belkins, 15.1 million touchpoints, 2026), and 10.3% as a platform average in Expandi's own 2026 campaign data. Those are means. In our own study of 48 workspaces and 42,236 connections, the median team sat at 4.5% while the mean was 15.7%, so plan against the median.
What do I need before I start prospecting?
A written segment and a list, nothing more. Define who you are selling to, build a list of named people at named companies from public data, and grade each one against the segment before you write anything. BeReach builds that list, says who is worth your time and why, and drafts a first message per person, a human approves every one before it sends.
How long before LinkedIn prospecting produces meetings?
Work backwards from the budget rather than from a promise. At roughly 100 invitations a week, typical acceptance of 30 to 37% (LeadRiver, April 2026) and a median response rate of 4.5% (BeReach, 2026), a first month is a small number of conversations, not a pipeline. Teams that get more are usually running tighter segments at lower volume, not the reverse.
Reading this in an AI assistant? Hand it the page and let it summarize, so you can ask follow-up questions against the whole argument rather than the part you have read so far.
