Most people qualify a LinkedIn lead in the wrong order. They find a name, send the invite, wait, and only start asking whether this person was a fit once the conversation is going nowhere. By then the invite is spent, the account is a little closer to its ceiling, and the answer arrives too late to be useful.
Turn the order around and almost nothing changes about the work, but a lot changes about the cost. The sorting happens first, on information that is already public, so the one scarce action you have each day goes only to people who already passed everything else.
What each check costs is the part that decides whether the method survives contact with a real send budget. Plenty of guides tell you what to check. Almost none of them tells you what each check costs.
Qualification is cheap, sending is not
Reading public information about a prospect is free and effectively unlimited. Sending an invite, visiting a profile or sending a message is metered, by the platform and by any tool that paces its sends. That asymmetry decides the whole method: run the free checks exhaustively, and spend the metered actions only on people who survived them.
Plan at about 100 invitations a week per account, and treat that as a budget, not a target. BeReach finds the right people and drafts each message, so the budget goes to the people who survived the free checks.
LinkedIn itself publishes no official ceiling for connection requests. The industry consensus sits around 100 a week, which works out to roughly 15 to 20 on a working day, and it is consensus rather than documentation. Read our full breakdown of the connection request limits for how the two layers interact.
The point is that a week's invitations are a finite, non-renewable thing. Spend one on somebody whose title you never checked and you do not get it back next week.
What public information can establish on its own
Role, seniority, employer, company size, industry, location, a recent job change, open roles the company is hiring for, published posts and public comments are all readable from public sources. For most B2B offers that set is enough to settle the fit question completely.
| What you want to know | Signal that answers it | When to check it |
|---|---|---|
| Do they hold the role I sell to? | Headline and current title on the public profile | Free check |
| Which company, and is it the right kind? | Employer on the profile, plus the public company page | Free check |
| How big is the company, and in what sector? | Public company page: size band, industry, headquarters | Free check |
| Are they investing in the function I serve? | Public job posts on the company page | Free check |
| Do they care about my topic right now? | Their comment under a relevant public post | Free check |
| What do they publish about? | Public posts found by keyword or hashtag | Free check |
| Did they change role recently? | Position dates on the public profile | Free check |
| Who reacted to a post without commenting? | Reactor and liker identity | Optional depth |
| Everything one person posted, in feed order | Their own post feed | Optional depth |
| Are they 1st, 2nd or 3rd degree to me? | Connection degree filters | Optional depth |
| Headcount bands and seniority facets | Sales Navigator | Optional depth |
| Have we already spoken? | Your inbox and connection list | Optional depth |
| Send the invite, send the message | The action itself | The spend |
Two rows deserve a closer look because people usually get them backwards. Commenters on a public post are public: the comment text is right there, and it tells you what the person thinks, not just that they scrolled past. Reactors and likers are not exposed the same way, so a reliable identity for them is extra depth rather than a free check. If you are choosing which one to build a qualification pass around, the commenters are both cheaper and richer. Our comments finder reads them straight from the public post, and the free people search finds people by role and place, with their employer.
What sits behind LinkedIn's own login, and what actually touches someone
Three things sit behind LinkedIn's own login: identifying who reacted to a post, reading one person's own feed of posts, and filtering by connection degree or through Sales Navigator. None of those three spend anything from your daily budget on their own. The profile visit, the invitation and the message are a different category: they touch the other person directly, which is what makes them the metered actions. Everything else is optional depth.
That list is shorter than the tooling market implies. Sales Navigator is the clearest example: it is genuinely useful for headcount bands, seniority facets and saved-lead alerts, and those are genuinely locked behind the subscription. LinkedIn's compare-plans page in August 2026 lists Core at $119.99 per month billed monthly, or $1,079.88 per year, which works out to about $89.99 per month, and Advanced at $159.99 per month billed monthly, or $1,799.88 per year [1]. If the filters you actually use are role, company and sector, you are paying that subscription for a lane you could have read publicly. There is a fuller treatment in intent signals you can track without Sales Navigator.
The profile visit is the subtle one, and it is where most qualification advice quietly contradicts itself. Guides tell you to visit the profile and study it before you decide. Reading a public profile is invisible and free. Visiting the profile itself puts you in the target's "who viewed your profile" list, which is a real touch, and it spends from the daily visit budget like any other outreach action. A research step that consumes send budget is not research, it is outreach with the ask removed. Treat it as outreach, because that is what it is.
Order the checks by cost, and stop at the first failure
Sequence qualification cheapest check first, and drop a lead the moment one fails. Done in that order, every disqualification is free, and the only leads that cost you an account action are the ones that already passed everything else. Most published checklists list the same checks in no particular order.
| Order | Check | What it removes | Cost |
|---|---|---|---|
| 1 | Start from a source with intent already in it: commenters on a post about your problem space, or companies hiring for the function you support | Everyone who has shown no sign of caring about the problem | Nothing |
| 2 | Current title and headline against the ideal customer profile | Wrong role, wrong seniority, wrong function. This kills the largest share of any list | Nothing |
| 3 | Employer: sector, size band, location, from the public company page | Right person, wrong company. A perfect title at a company you cannot serve is still a no | Nothing |
| 4 | A reason for the message to arrive this week: recent role change, an open req, a comment from the last few days | People who fit but give the message no reason to arrive today | Nothing |
| 5 | Write the first line. If it could have been sent to anyone else, go back to step one | Leads you only thought you had qualified | Nothing |
| 6 | Your own inbox and connection list, to check a colleague did not already reach out | Duplicates and awkward second approaches | Nothing spent |
| 7 | Visit, invite or message | Nothing. This is the spend | One action against a hard daily ceiling |
Steps one through five run on public information alone. BeReach builds the list for you with full details on every person, finds the dated reason for step four (a new role, a recent post, a funding round or an open job), drafts a first line worth sending, and a human approves every message before it goes. For the same idea applied to ongoing signals rather than one-off lists, read signal-based selling on LinkedIn.
Step four is the one people skip, and it is worth naming why it belongs in the qualification pass rather than in the copywriting pass. "No reason to arrive today" is most of what makes outreach feel cold. A lead with no trigger is not disqualified forever, it is disqualified this week, which is a different verdict and should send them back to the pool rather than to the bin. Hiring signals are the most reliable public trigger for the same reason they are the most public: a company that has posted three roles in your function has told the internet where its budget went.
Public data has a shelf life
Public profile data goes stale. Front-loading qualification works, as long as the front-loading happened recently.
Use gates before you use a score
Qualify with boolean gates, and add a weighted score only once you have enough closed-won data to fit the weights against. Additive scoring lets a lead collect soft points until it passes while failing something that should have been fatal. Gates cannot do that, and they cost less to run.
Additive scoring models for LinkedIn outbound tend to have the same shape. What follows is an illustrative composite of that shape, written out so the arithmetic below is checkable, not a surveyed distribution: 0 to 25 points for title match, 0 to 20 for company size, 0 to 20 for industry, 0 to 15 for seniority, 0 to 10 for geography, then an intent layer on top. The shape has two problems and neither is usually stated.
The first is that the weights are guesses unless a closed-won dataset produced them. A model fitted to nothing is your existing prior with a decimal point attached, and the decimal point is the dangerous part, because it makes a hunch look like a measurement. Fitting the weights needs enough won and lost deals to separate signal from noise, which most teams starting LinkedIn outbound do not have yet, and will not have for several quarters.
The second is that addition hides vetoes. On the composite above, a lead with a perfect title, perfect industry and perfect geography banks 55 of the 100 before anyone looks at anything else. If they run a two-person consultancy and you only sell to companies above 200 seats, that is not a 55, it is a no. Additive models turn hard disqualifiers into soft deductions, and the leads that survive on accumulated near-misses are exactly the ones that waste invitations, because they look plausible on the row and fail on the call.
The fix is ordering rather than abolition. Gate for inclusion, in the cost order above, where each check is a yes or a no and a no ends the evaluation. Score only among the survivors, and only to decide who gets contacted first when 400 people passed and you have about 100 invitations this week. That is the job a score is actually good at, and it is a much smaller job than the models imply. Our guide to building a B2B lead scoring model covers the prioritisation half in more depth.
Describe who you want.
Get the people.
The people search opens one person from a name, a name and company, or a profile link, and finds many by role, company and place, in plain words. Free to try.
Search people
Why front‑loading qualification lifts acceptance
Qualifying first raises acceptance through one mechanism: knowing something specific before you contact someone. A generic note at volume reads like a template. A specific note reads like a person who did the work, and the acceptance rate that follows reflects that difference.
LeadRiver's April 2026 analysis of more than 50,000 connection requests puts typical B2B outbound acceptance at 30 to 37 percent, and finds that trigger-based notes, meaning notes that reference something specific about the person, were accepted at 50 to 60 percent against 15 to 25 percent for generic ones. You cannot write a trigger-based note about someone you have not qualified. The personalisation and the qualification are the same act performed once, which is also why step five of the cascade is a qualification check and not a writing tip.
Run the arithmetic on a week's budget. Take the 100 invitations a week you plan around. At a generic 20 percent acceptance you finish the week with 20 accepted connections, and if half the list was never a fit, 10 of them are useful. Fill the same 100 slots with qualified people and trigger-based notes, and at 50 percent acceptance you finish with 50 relevant connections. Same account, same budget, same week, five times the useful output.
Here is the part the case-study version leaves out. To fill 100 slots at a realistic pass rate you have to screen several hundred people. At 90 seconds of human attention per profile, screening 250 takes over six hours, most of a working day spent every week before a single invitation goes out. Hand-qualifying to fill a week of invitations does not pay for itself at any sensible cost of an hour. The ratio only works when steps one through four run without a person reading each profile, and a human attention budget is reserved for step five and the approval before the send. That is the real case for tooling here, and it is a narrower case than the market makes.
Relationship state compounds the acceptance effect. Belkins' 2026 study of 15.1 million touchpoints found 7.9 percent replies on a cold connect-then-message sequence against 12.2 percent when messaging someone already connected. That compares warm against cold rather than qualified against unqualified, but it points the same way: the further you get before the ask, the better the ask performs. For platform-level context, Expandi's 2026 benchmarks put the average LinkedIn reply rate at 10.3 percent against 5.1 percent for cold email. 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.
The volume argument explains the acceptance numbers, not virtue. Sending carefully at 20 a day and sending indiscriminately at the ceiling draw down the same weekly budget, but indiscriminate sending spends most of it on people who were never going to reply. A string of withdrawn invites and ignored requests is what a poor acceptance rate looks like at volume, and qualifying first is what keeps that ratio in the useful range.
Three qualification numbers that do not survive checking
Three figures turn up in nearly every guide you will read, and all three are attributed to the wrong publisher. The findings themselves are mostly real. What has been lost is who measured them, what they measured, and whether the measurement had anything to do with LinkedIn outbound at all.
| How it circulates | What the original actually says | Who actually published it |
|---|---|---|
| "LinkedIn allows about 100 connection invitations a week", often cited to the LinkedIn Help Center | LinkedIn documents no invitation ceiling anywhere. There is no published figure to cite | Practitioner consensus, repeated until it read as documentation. Useful as a working assumption, wrong as a citation |
| "Replying within five minutes lifts conversion 8 to 10 times", usually credited to a Salesforce report | The odds of qualifying a lead drop 21 times between a five-minute and a thirty-minute callback, and the odds of contacting drop 100 times | Dr James Oldroyd, MIT Sloan with InsideSales.com, 2007. Six companies, 15,000+ leads, 100,000+ call attempts, and it measures inbound web leads, not outbound |
| "B2B contact data decays about 23 percent a year", sometimes credited to LinkedIn's Economic Graph | B2B data decays about 2.1 percent per month, which annualises to 22.5 percent | MarketingSherpa, republished on HubSpot's database decay page, which gives no date for the underlying research [2]. It measures email contact databases, not LinkedIn profiles |
None of this makes the underlying research bad. It makes the citations useless, and in the third case it makes the number misleading twice over: a decay rate measured on email addresses tells you nothing about how fast a LinkedIn headline goes stale, and nobody in the republishing chain will tell you what year it was measured.
If you want a figure for the speed penalty that is correctly sourced, use the follow-on work, and keep its two datasets apart, because they get fused constantly. Oldroyd, McElheran and Elkington in Harvard Business Review, March 2011 audited 2,241 US companies with a test web lead, and that audit produced one thing: a response-time distribution, in which 23 percent of the companies never responded at all [3].
The seven-times and sixty-times multipliers come from a different dataset, described in the same article as a separate study of 1.25 million sales leads received by 29 B2C and 13 B2B US companies. There, firms that made contact within an hour were nearly seven times as likely to qualify a lead as those that tried an hour later, and more than sixty times as likely as those that waited 24 hours or longer [3]. Both are inbound, and both should be labelled inbound whenever they are used.
The habit worth taking from this is small and cheap. Before you repeat a number in a sales deck, open the link. If the link goes to a blog post that cites another blog post, you do not have a statistic, you have a rumour with a percentage sign.
What public qualification cannot tell you, and how fast it expires
Public data settles fit. It does not settle budget, authority, timing or intent, and a qualified list is perishable rather than permanent. Titles change, companies restructure, and the reason you had for messaging someone in March is not a reason in July. Know both edges before you trust a list.
On the first edge, you will not get a verified work email, an exact headcount, a confirmed reporting line, or any signal that this person is actually shopping. That is why the message still has to ask a question rather than assume an answer, and why a well-qualified list can still produce a bad quarter if the offer is wrong.
On the second edge, treat the qualification like inventory rather than an asset. Fit signals age slowly: an industry rarely changes, a company size band moves over years. Trigger signals age fast: a job post is stale in a month, a comment is stale in a fortnight, a role change stops being news after a quarter. A list qualified in one pass and worked over six weeks is really two lists, one still valid and one that quietly expired, and nothing in the list itself will tell you which row is which. That is the argument for a refresh rule rather than a refresh habit: decide how old a signal may be before a row is checked again, and write it down.
The one check that genuinely depends on your own inbox is whether a colleague already contacted this person, and it is worth doing last precisely because everything before it can be settled from public information alone.
So public qualification removes the people who were never going to be a fit, cheaply and at scale. It cannot manufacture a buyer, and it cannot stop a good list going off. What it buys you is that every metered action you spend goes to someone who at least deserved it.
Frequently asked questions
How much of LinkedIn lead qualification is based on public information?
Most of it, for most B2B offers. Role, seniority, employer, company size, industry, location, recent role changes, open job posts and public comments are all readable from public sources, and that set answers the fit question on its own. Reactor identity, a person's own post feed, degree filters and Sales Navigator sit behind LinkedIn's own login and add depth rather than a verdict, and your own inbox and every outreach action are a separate, later step.
How much profile data do you actually need to qualify a lead?
Less than most people assume.
Should I use a lead scoring model or a simple checklist?
A checklist of boolean gates first, and a score only for ranking the survivors. Additive scoring lets a lead accumulate soft points and pass while failing a criterion that should have been fatal, and the point weights are guesses until you have closed-won data to fit them against. Gate for inclusion, score for the order you contact people in.
Does visiting a profile to research a lead count against my limits?
Reading a public profile does not. Visiting the profile itself does, on both counts: it consumes daily visit budget and it shows up in the target's "who viewed your profile" list, which makes it a touch rather than research.
How many leads should I qualify per day?
Work backwards from the send budget. Plan at about 100 a week per account, roughly 20 a working day. Screening a few hundred people a week to fill those invitations well is the right ratio. The catch is time: screening several hundred profiles by hand costs most of a working day, so the first checks have to run without a person reading each one.
Qualify first, send second. BeReach builds the prospect list, adds a dated reason to write to each person now, drafts a first message per person, and a human approves every one before it sends, in the app or inside Claude. BeReach is EUR 99 per month billed monthly on Pro, per workspace rather than per seat, with a 14-day free trial. See how BeReach runs your outreach.
Sources and references
External sources cited in this article. Each numbered mark points to the source behind that passage.
- LinkedIn: Sales Navigator compare plans
Core and Advanced monthly and annual prices, as listed in August 2026.
- HubSpot: Database decay
The MarketingSherpa finding of about 2.1 percent B2B data decay per month.
- Harvard Business Review: The Short Life of Online Sales Leads
The 2,241-company response audit and the seven-times and sixty-times contact multipliers.
