First the connection. Then the conversation.

Early customers aren’t buying a finished product. They’re betting on you. Max Kolysh, a YC Visiting Partner who asked dozens of YC founders how they found their first ten customers, says that if someone is going to bet on a new product from a new company, “they at least need to trust the person building it.” ¹ LinkedIn is where that trust gets checked. Before anyone answers you, they can see who you are, what you’ve done and who you both know.

So use LinkedIn to meet people, not to pitch them. Zip, a YC company from summer 2020, got its first 10 customers “by reaching out for advice on LinkedIn,” and later in this guide you’ll see how it turned that advice into sales. ⁸

It isn’t the best first channel for everyone. Aaron Epstein, a YC General Partner and Creative Market’s co-founder, finds that “email is usually better, LinkedIn is usually worse,” though it depends on the industry: people flooded with LinkedIn pitches tune them out, while people who rarely get one find it special. ⁹ And many buyers don’t sit at a laptop at all. ¹

In a survey that ran alongside a LinkedIn field experiment by three economists, users said they spend around three minutes deciding on a connection request, and the red flags they named most were suspicious profile entries, profile pictures and posts. ² A profile that reads like a sales page turns your blank invite into a pitch before you’ve said a word.

Here’s what I’d fix before sending a single invite. (Mosswood, in the headline example, is the made-up startup used in this guide’s sample messages. It cleans up the CRM, the software where a sales team tracks its customers.)

• A real photo. Carl Cortright, a YC founder (Flyflow, S24), puts it simply: “keep it to a smiling headshot, shoulders and above that clearly shows who you are.” ¹⁰
• A plain headline. Your role, your company and what you’re building for whom: “Co-founder at Mosswood · cleaning up CRM data for Series A sales teams.” Add YC or a known investor if you have one.
• A short About, in your own voice. Three or four sentences on what you’re building, why, and what you did before. No call to action.
• Featured shows your work. Pin a short demo, your launch post or something you wrote about the problem. Not a booking link.
• A quiet banner. Your product’s name and one line, or a plain image. “Book a demo” tells people a pitch is coming.
• Signs of life. Post now and then about what you’re learning or shipping. Users in the same survey flagged both a lack of activity and excessive posting. ²
• People you know, first. Connect with former colleagues, classmates, investors and other founders before any stranger. They’re your warm paths: one YC founder got intros to about half the customers she closed during her batch through LinkedIn. ¹

Start with people who already trust you. Kolysh’s order is your own network, then your second-degree connections (ask the person in the middle for an intro), and only then strangers: “If you haven’t worked your network, you’re skipping the lowest hanging fruit.” He also finds that “the exact copy matters less than people think.” Who you write to does more. ¹ The prospecting guide covers building the full list, with signals and emails.

Plan on hundreds of people, not dozens. Gustaf Alströmer, a YC General Partner, sketches a funnel where 500 emails or LinkedIn messages turn into about 10 calls and 2 customers. It’s an illustration, not a benchmark, but it sets the right scale. ¹² The cold outreach overview works it backward from the number of customers you want.

He adds that one recent YC founder got his biggest customer this way, and that it was his highest-converting cold outreach channel. ¹

Why a blank invite works:

• They’ve already said yes to you once. Accepting is a small decision about you, made after looking at your profile. The economists behind the LinkedIn experiment suggest people then “stick to past decisions and treat connections in their network equally.” ²
• You’re in their network. In the same study’s survey, being in someone’s network was the biggest factor in whether they’d answer a message, other than knowing the person. ²
• It’s normal. Only about 22% of LinkedIn users surveyed said requests come with a message at least most of the time. ² And blank invites get accepted at least as often as ones with a note: Belkins, a lead-generation agency, saw 27.6% against 25.3% across its own campaigns. ¹³
• A note reads as a pitch. Cortright finds that “anything but a plain connection request (no message attached), comes off as too salesy.” ¹⁰ LinkedIn’s own policy: “Do not use our invitation feature to send promotional messages to people you don’t know.” ¹⁴
• You couldn’t write many anyway. A free account can add a note to three invites a month, 200 characters each, as of October 2026. ¹⁵

My rule: add a note only when you share something real. A school, a past employer, a talk of theirs you were at, a friend in common. Make it specific, keep it under 200 characters so it fits a free account, and leave the pitch out.

Send in the morning, at a steady pace. Expandi’s data from its own customers puts acceptance at about 32% for invites sent between 7 and 11 a.m., against 24% in the evening, with the day of the week barely mattering. ¹⁶ Fifteen to twenty invites a weekday keeps you near 100 a week, about where accounts commonly hit LinkedIn’s limit as of October 2026 (more on limits below). Withdraw invites still pending after two or three weeks; once you do, you can’t invite that person again for up to three weeks. ¹⁸

Kolysh’s rules for the message itself are short: “keep it under 75 words. Long emails tend to just get ignored, and they’ll assume you used an LLM to write it.” (An LLM is an AI writing model like ChatGPT.) And “the single most important line is a clear call to action.” ¹ On LinkedIn I’d go shorter still: two to five sentences. Make it obvious why you picked this person, and use Pete Koomen’s test: “only send emails that you yourself would be excited to read.” ¹¹

Start a conversation, not a sale. As Will Allred, Lavender’s co-founder, puts it: “You’re not pitching. You’re starting a conversation.” ⁵ Kolysh found that for many founders “the most effective early outreach wasn’t actually framed as a sales pitch,” but as a request for advice or mentorship, a product review or a whiteboard session. ¹ Legora’s founders, building legal AI, wrote to lawyers on LinkedIn asking for a lunch to learn about their practice areas, and offered to pay their hourly fee. ¹⁹ Asking doesn’t make you look weak, either: in a 2015 study, people rated those who asked for advice as more competent than those who didn’t. ²⁰ Here are three honest ways to open.

A real one. Aaron Epstein says he answers very few LinkedIn messages, but he answered this one. It opened with “Hope you don’t mind the cold reach out, but saw you also spent time in Van Munching Hall,” a building where he took classes at the University of Maryland. One line said who they were. Then came the favor: “I was hoping you might help out a fellow Terp and review my YC app.” It ended with an easy next step: “let me know the best email to send you the doc.” ⁹

Three rules for any of them:

• Don’t disguise a sales call. If you ask for advice, mean it. Kolysh: “If you’re not actually open to learning from them and you’re just trying to disguise a sales call, don’t do this.” If you’re selling, say so and ask straight for 15 minutes. ¹
• Ask about their world, not your idea. Rob Fitzpatrick’s The Mom Test sums it up: “Talk about their life instead of your idea.” How they handle the problem today tells you far more than what they think of your pitch. ²¹
• Only ask people who have the problem. Koomen warns that feedback from people who don’t need your product “is useless at best and counterproductive at worst.” ¹¹

One message is often not enough. As Epstein says, “People are busy. They’re out on vacation. They get distracted.” ⁹ Kolysh’s rule of thumb for early outreach is to follow up “three or four times over a couple weeks.” ¹ On LinkedIn I’d stop at two or three messages and let email carry the rest. Will Allred caps LinkedIn at three touches and moves the third to email. ⁵

The timing I’d use: your first message the day they accept, a follow-up four or five days later, and about a week after that, an email plus a one-line note on LinkedIn pointing to it. Then stop. Expandi’s data from its own customers points the same way: “Three messages is the campaign sweet spot,” and five or more did worse than sending just one. ¹⁶

If they still don’t answer, let it go. Epstein again: “Nobody owes you anything here. Just move on.” Check back in a couple of months, when the problem may matter more to them. ⁹

Ask for 10–15 minutes by phone or Zoom, and say so in the message. Kolysh: “Are you offering a 15-minute demo? Make it really clear. Otherwise, they’ll assume the worst.” ¹ Gustaf Alströmer’s sample LinkedIn message to a former colleague is just as plain: “I just ask for a 20-minute phone call or video call. That’s it.” ²² I’d go shorter. Fifteen minutes is easy to say yes to, and you can keep talking if it’s going well.

When they say yes, offer two times instead of a booking link. Nick Abraham, who founded the cold email agency Leadbird, is blunt about it: “Calendly links can make it harder to book cold leads.” ²³

• Say what you’re working on in one sentence, then ask your first question.
• Ask about the last time the problem came up and what they did. The Mom Test calls this asking about “specifics in the past instead of generics or opinions about the future.” ²¹
• Ask what they use today, what it costs them and who else deals with it.
• Talk less. Tyler Bosmeny, Clever’s co-founder, says the number one thing to remember is “when you get them on the phone, remember to shut up.” ²⁴
• Find out whether they have the problem, the budget and the say to buy. That’s Koomen’s goal for a first call. ¹¹
• Show the product only if they ask, or book a second call for it.
• End with the next step on the calendar. Tom Blomfield, a YC General Partner and Monzo’s co-founder: “Don’t leave a meeting without the next touch point set up.” ²⁵

This is how Zip did it. Rujul Zaparde, its co-founder and CEO, says they “would max out our LinkedIn connections diligently every morning,” then messaged the people who accepted and asked, truthfully, for advice. Once they had built what people needed, they went back with “hey remember that conversation we had your feedback was so helpful to us,” then “we built it can we show it to you.” That’s how they closed their first customers. ⁴

Kolysh tells a similar story about the founder who tested one guess a week: “By the time she had a product, she already had a pipeline of people who told her exactly what they needed.” ¹ That’s know, like and trust doing its work: people watched you listen, then saw their own idea in your product.

1. 1

   Keep a row for everyone you talk to.

   What they said, what they need and when to come back. A spreadsheet is enough to start.

2. 2

   Go back when you ship what they asked for.

   Name their idea, show it working and ask for 10 minutes. One person at a time, not a newsletter.

3. 3

   Ask for money early.

   Gustaf Alströmer: “if you don’t charge your customers, they are not a customer, and you don’t have a company.” ¹²

A relationship isn’t a reason to stall, though. Blomfield: “Honestly, most founders are too slow to progress through these stages.” ²⁵