You already have 20,000 connections

Twenty thousand, two thousand, four hundred. Whatever the number on your profile is, you are paying for a list of strangers while sitting on a list of people who already accepted an invitation from you.

That is a familiar argument and usually a soft one, made out of sentiment about relationships. There is a harder version, and it comes from LinkedIn's own pricing behaviour: LinkedIn charges you nothing to work the network you have, and meters almost everything you do to find people you do not.

Two separate mechanisms, both stated on LinkedIn's own help pages.

What LinkedIn charges you for

LinkedIn runs a monthly budget called the commercial use limit. When your activity looks like hiring or prospecting, you hit a ceiling on searching and viewing profiles.

The page describing it lists exactly what spends the budget:

• Searching for profiles on LinkedIn.com and mobile
• Browsing profiles through the People Also Viewed panel
• Viewing member profiles on the People tab of a LinkedIn Page

And what does not:

• Searching profiles by name in the box at the top of the page
• Browsing your first-degree connections from the Connections page
• Searching for jobs

Read the middle item again. Working through the people you are already connected to costs nothing against a budget that a filtered search spends.

The budget itself is invisible. LinkedIn says it is "not able to display the exact number of searches or views you have left and we also cannot lift the limit upon request," and that the warning you are approaching it "may not display if you run through the full amount of searches or views too quickly." Upgrading does not remove it either: LinkedIn's own wording is that a Premium plan will "increase the number of profile searches and views." We took that apart in what counts against your LinkedIn search limit.

One honest qualification. LinkedIn writes the exemption against a place, the Connections page, rather than against first-degree status in general. Nothing published says what happens when a first-degree connection turns up inside a filtered search, and it would be reasonable to assume that search still costs you. Work the list from the Connections page.

The second mechanism, and it is the sharper one

Messaging follows the same shape.

LinkedIn's help centre: "On LinkedIn, you can message your 1st-degree connections or LinkedIn Pages for free." Reaching someone you are not connected to normally needs an InMail, InMail is a paid feature, and LinkedIn states plainly that "We currently don't offer a subscription plan with unlimited InMail message credits." The free exceptions LinkedIn lists are narrow and mostly outside your control: the recipient is a Premium member who has turned on Open Profile, you are both in the same LinkedIn Group, or you are current coworkers.

Then there is the detail that gives the whole thing away. LinkedIn's own wording is that you "can receive InMail message credits back if a recipient accepts your InMail message." The platform charges for contact and gives the charge back when the recipient engages with it.

So the two prices LinkedIn puts on reaching a person are zero and metered, and the line between them is whether that person already agreed to hear from you.

If you have a few thousand connections and you have never worked them, you have been buying the expensive side of that line while the cheap side sat there.

The honest part: nobody has measured this

You would expect a number here. How much better does a first-degree message perform than a cold one?

It does not exist in citable form, and we looked.

The largest LinkedIn outreach dataset in circulation, Expandi's analysis of more than 70,130 campaigns published in January 2026, reports aggregate connection and response rates with no breakdown by connection degree at all. The figures people quote for warm versus cold do not survive being opened. The most-cited one, a 34% warm reply rate against 5% cold, comes from a lead generation vendor surveying its own audience, and the vendor's own body text says something different from its headline: 10% to 34% for warm, 2% to 10% for cold. A range of 10 to 34 is not a finding, it is the absence of one. The other line in circulation, that warm introductions convert ten to twenty times better, is attributed on the page carrying it to nothing at all. What that page does cite, a Harvard Business School survey of nearly 900 institutional venture capitalists, is about where venture deals originate, which is a different referent, a different metric and a different population from a founder sending a message.

The closest thing to a rigorous study on tie strength complicates the picture rather than supporting it. Rajkumar and colleagues, publishing in Science in September 2022, ran randomised experiments across more than 20 million LinkedIn members over five years, covering 2 billion new ties and 600,000 new jobs. They found that weaker ties increased job transmission and the strongest ties transmitted jobs least, with the effect flipping by industry.

That measures job mobility rather than sales replies, so it does not settle the question either, and the argument in this article never needed it to. The case rests on what LinkedIn charges, which is documented above and is not in dispute. What the study does suggest is that the useful people in your network are probably not the ones you talk to. They are the ones you barely remember connecting with, which is exactly the group nobody works.

So: no measured multiple, a documented price difference, and reasoning from it. Anyone quoting you a warm-versus-cold reply-rate multiple is quoting something we could not find measured anywhere, in academic work, vendor benchmarks or platform reports.

What is actually in the file

LinkedIn will hand you the list. Settings, Data Privacy, Download your data. Do not plan around getting it in minutes: LinkedIn sorts its data categories into one table headed "Available within 10 minutes" and another headed "Available within 48 hours", and Connections sits in the 48-hour table. The same page's prose says the larger archive arrives "within 24 hours", so LinkedIn publishes two different waits on one page. The download link expires after 72 hours, and the feature does not work on mobile.

What comes back, in LinkedIn's own description of the Connections file: "the first and last name, public profile URL, email address, company, position, and connection date for all 1st degree connections."

Seven columns, and one of them is conditional. LinkedIn's wording: "You may notice in your connection data that some of your connections' email addresses are missing. That's because members choose in their privacy settings whether to allow others to download their email address."

That is the expectation gap. People run the export believing they are getting twenty thousand email addresses and get names, companies, job titles and dates, with emails only where the other person allowed it. We walked through the whole process in how to export your LinkedIn connections.

Worth knowing before you plan around the number in your profile: LinkedIn caps first-degree connections at 30,000.

The problem is not the list. It is that the list is out of date

Look at what those columns actually are.

A first and last name, a profile URL, a company, a job title, the date you connected, and an email address where the person allowed it. The company and the title are as they were when LinkedIn last knew, and the date tells you how long ago you had a reason to talk. Nothing in that file tells you whether the person still works there, whether they moved into a role that could buy from you, whether the company grew, or whether anything happened that makes this a reasonable week to write.

A connection from 2019 is a person whose current situation you know nothing about, which is the same problem a cold list has. What is different is documented rather than assumed: you can already reach them, at no cost, without spending a search or a credit.

That is the actual work: not finding these people, which is done, but finding out who they are now. Take the export, attach the current context to it, and the list you already own becomes a list you can write to.

Request the export today, because Connections sits in LinkedIn's 48-hour band and nothing starts until the file arrives. Then, against the finding above, do not start with the twenty people you already talk to. Sort the file by connection date, take twenty from three or four years ago whose company or title you cannot recall, and look up what those twenty are doing now. If none of them is worth a message, the network was not the opportunity and you have learned that for the cost of an afternoon.

LinkedIn's help centre pages on the commercial use limit, messaging, InMail, data download and network degrees were read on 3 September 2026. The Expandi and Science figures are cited to their own publications.