How to Find Customers for a Startup: Your First 10 Buyers
Where a startup's first buyers show up on LinkedIn once the warm intros are spent, the exact searches and messages for each signal, a week-one plan one founder can run by hand, and when referrals, communities or paid beat signals.
The fastest way to find customers for a startup is to sell to the people you already know, then switch to people who are describing your problem in public before your network runs dry. For a B2B startup that second group lives on LinkedIn: founders complaining about the thing you fix, people asking which tool to use, commenters under a competitor's post. This page covers where those buyers show up, a week-one plan you can run by hand, and when referrals, communities or ads are the better bet.
Why does the usual advice stop at customer five?
The pages that rank for this query agree on the order: write down an ideal customer profile, sell to your own network, ask for referrals, then "do outbound" and "start marketing". The first three steps work and you should do them. They also run out, usually somewhere between the fifth and the tenth customer, because a founder's network holds only so many people with the exact problem and the budget to fix it this quarter.
What the pages skip is the next step: how to find strangers who have the problem now. "Do outbound" without a way to pick who to contact means buying a list and messaging everyone on it, which is how a new company with no brand earns almost no replies and a restricted LinkedIn account.
Paul Graham's advice in Do Things That Don't Scale is to recruit users by hand at the start, one at a time, the way Stripe's founders set up beta users on the spot and Airbnb's founders went door to door in New York; his arithmetic is that 10% weekly growth from 100 users reaches about 14,000 in a year. The question this page answers is which doors to knock on. The answer is the people already saying, in public, that they have the problem. The buying signal lesson defines that behaviour; the rest of this page is where to find it when nobody has heard of you yet.
Where do a startup's first customers show up on LinkedIn?
A startup with no brand has one advantage over an established vendor: nothing to protect, so you can answer a stranger's question in full, including where your product does not fit. The buyers worth finding are the ones who make their problem visible. The 2024 Buyer Experience Report from 6sense, summarised in its press release, found B2B buyers are nearly 70% through their purchase before they engage a seller and 81% already have a preferred vendor at first contact. A startup that waits for inbound is the vendor who arrives after that choice is made. The way in is the research stage, which is where people ask questions in public; in G2's 2024 survey of more than 1,900 software buyers, research was the stage at which most buyers (34%) said a salesperson's contact was useful.
Each row below is something you can see with a free LinkedIn account. The last column is the message, written to quote the signal rather than pitch the company. Reddit and X carry the same questions and are worth a manual search a few times a week; treat them as manual channels, because Intentsly monitors LinkedIn only.
| Signal | Where | What it means | What to do |
|---|---|---|---|
| A post describing the problem you solve, in the words a customer would use | Posts tab, search the pain phrase in quotes, date filter past week | A problem acknowledged in public, no vendor chosen | Comment with a method they can use without you, then connect with a note that quotes the post |
| "What are you using for [job]?" or "[incumbent] alternative?" | Posts tab, search the incumbent's name plus "anyone using", "recommend", "alternative", "switching" | An active comparison, often with a deadline | Answer in the thread first, including where you are the wrong choice; message with the one reason you fit the job they named |
| Comments and reactions under a competitor's or an industry leader's post | The competitor's company page, its founders' posts, the three or four people who write about your category | The person follows the category closely enough to react; a question in the comment is an open evaluation | Reply to their comment within a day, then connect with a note that refers to it |
| A new role in a buying title ("excited to join", "new role", "first 90 days") | Posts tab, search the phrase plus the title you sell to | New owners audit the stack and the suppliers in their first quarter | Congratulate, ask what they inherited, offer a comparison note rather than a demo |
| A founder posting about the same stage you are at (first hire in the function, first sales process, first tooling budget) | Posts tab, search "first sales hire", "first SDR", "setting up our", plus your category word | Budget is being created for the outcome you sell; nothing is in place yet | Share what you did at that stage, then ask what they are considering |
| Someone in your target title newly follows a competitor page or engages with the category hashtag | The competitor page's follower list; posts tagged with the category hashtag | Early research, weaker than a comment but fresh | Connect with no pitch and let your posts do the next step |
One filter applies to every row: a signal older than about two weeks is cold. The person who asked "anyone using X?" a month ago has usually chosen, and the founder who posted about hiring a first SDR has usually hired one.
How to find customers for a startup by hand: a week-one plan
This plan takes about an hour a day, a free LinkedIn account and a spreadsheet. It assumes you have already sold to the people you know and taken their referrals. The company in the example is hypothetical.
- Day 1: write the signal list. Three incumbents or substitutes your buyers use today (for a new category, the spreadsheet or the agency counts), five phrases customers used for the problem on your first calls, the two or three titles that bought. Example (hypothetical): Harborline, a new tool for agency invoicing, lists two invoicing tools and "spreadsheet" as incumbents; its phrases are "chasing invoices", "net 30 is killing us" and "which invoicing tool"; its titles are agency founder and operations lead.
- Day 1: save the searches. In LinkedIn's Posts tab, search each incumbent name in quotes plus one of "anyone using", "alternative", "switching to", and each pain phrase in quotes. Set the date filter to the past week and save every search. Follow each incumbent's page and the three or four people who write most about your category, so their posts and the comments under them land in your feed.
- Day 2: build the watch list. Open the newest ten posts from your saved searches and the three most recent posts from each incumbent. List every person whose title matches a buying title and anyone who asked a question. Record name, company, post, the quote and the date. The decision-maker guide covers finding the right person at a company once you know the company.
- Days 2 to 5: act within a day. Reply in the thread first, then send a connection note that quotes what they wrote and asks one question. No link, no deck, no case study attached, because you do not have one yet and they know it. Keep to 15 connection requests and 15 messages a day; a restricted account ends the plan. The message templates for buying signals give one opener per signal.
- Days 3 to 5: answer the open questions in full. For every "what are you using?" post, write the answer you would give a friend, naming the incumbents that fit better for some cases. A startup that recommends a competitor where the competitor is right gets remembered when it is wrong.
- Day 5: score the week. Count signals found, messages sent, replies and calls booked. Replies under one in ten means the note is pitching; rewrite it to lead with their quote. Fewer than ten signals a week means the phrases are too narrow or the category is quiet on LinkedIn, in which case the alternatives below matter more.
- Day 6: ask every conversation for two names. Everyone who takes a call, buyer or not, knows two people with the same problem. This is how the warm network that ran out at customer five gets rebuilt from strangers.
Expect week one to produce conversations and a handful of calls, not signed contracts. The point is a loop you can repeat every week: watch, quote, reply, book.
How does Intentsly do this for you?
The manual plan breaks around week three. The saved searches must be re-run daily, competitor threads grow faster than one founder can read, and with sending capped the bottleneck is finding the right fifteen people each morning, not messaging them. Intentsly runs the finding and the first conversations, on LinkedIn only.
| Signal from the table above | What Intentsly does |
|---|---|
| Pain-phrase posts, "anyone using X?" posts, "first sales hire" posts | Keyword post monitoring for the pain points, tools and topics you set |
| Comments and reactions on competitor and industry-leader posts | Post engagement monitoring: people who like, comment or share posts from the competitors and leaders you name |
| New competitor page followers and engagers | Competitor page follower and engager tracking |
| Hashtag engagement in your category | Hashtag engagement monitoring |
| Deciding who is a buyer | Scoring against your ideal customer profile, so the list arrives prioritised |
| Finding an email for the follow-up | Enrichment with verified emails, and export of leads and signals |
| Replying, holding the conversation, booking | The AI SDR sends LinkedIn connection requests and messages, holds the conversation, and sends your Cal.com or Google Calendar link so the buyer books the meeting; HubSpot is the native CRM integration |
If you want the watching done for you from day one, the LinkedIn signals feature runs the saved searches and competitor-post checks from the plan above and hands you the list already scored. The limits are the same as the manual plan, stated plainly: up to 2 LinkedIn accounts per plan, a maximum of 30 messages and 30 connection requests per account per day, with 15 recommended. Monitoring and outreach run in English, Romanian, German and Dutch. The AI SDR sends LinkedIn messages only; it does not send email.
What does this look like in practice?
We were the startup in this situation. Intentsly launched an AI sales platform into a market full of outreach tools with bigger budgets and existing brand awareness, which is the position most readers of this page are in. We pointed our own intent signals at founders complaining about cold outreach burnout, the problem we had built the product around, and let the AI SDR handle the replies and the booking. The result was 14 qualified demos in 14 days with zero manual outreach, and a reply rate 4x our previous cold campaigns. Most of those leads came from signals caught in real time: the message arrived while the person was still describing the problem. The write-up is on the case studies page.
The part of that run that carries over to any startup is the targeting: we watched for one specific complaint, stated the way the people with it state it, rather than for a title or a company size. Build your watch list the same way, from the phrases your first customers used.
Other ways to find customers for a startup, and when they beat signals
Signals are the fastest route from "network exhausted" to a weekly flow of conversations. They are not the only route, and the B2B SaaS version of this page goes deeper on the product-company side.
Referrals and partners beat signals while the first ten customers are still being won, and again when your product is set up by someone else. In the same G2 survey, 69% of software buyers said service providers and implementers are a factor in their decision, and 38% start considering them during research, so the consultants and agencies already serving your buyers are a lead source before your brand is. Referrals are also the only channel that gets better as you get customers. The ceiling is low and the pace is slow, which is why the signal plan starts the week the intros run out, not after.
Communities and forums beat signals when your category is new enough that nobody searches for it yet. A Slack group, a subreddit or a niche newsletter where your buyers already gather lets you answer questions for months before anyone types your category into LinkedIn. The cost is time, and the rule is the same as the signal plan: answer in full, recommend others where they fit, and never post a link until asked. The Reddit lead guide covers how to do that without getting banned.
Paid ads beat signals when you already know which sentence makes a buyer click and need volume this quarter. Before that, ads pay to learn what a week of signal conversations teaches for free: the problem phrases in your buyers' own words. A bought lead list is the one method signals replace outright; a list says who could buy, not who is buying this month.
The next step is to check whether your buyers talk about the problem in public at all, because the whole plan rests on that. Try the AI shows matching buyers for your product with no account, which answers the question in a few minutes. If they do, the week-one plan is ready to run, and the Buying Signal Playbook lists the 12 signals worth chasing and the 8 to ignore so the watch list starts tight.
FAQ
How do startups find their first customers?
Almost always through people they already know: former colleagues, investors' portfolios, the first users who gave feedback. That gets most startups to somewhere between five and ten customers. After that, the founders who keep growing switch to finding strangers who show the problem in public, on LinkedIn and in communities, and treat every call as a request for two more names.
How do I find customers for a startup with no network?
Start with the people describing your problem in public. Search LinkedIn posts for the phrases a customer would use, set the filter to the past week, and answer those posts with something useful before you send any message. A stranger who got a good answer from you behaves like a warm contact. An hour a day with a free account is enough for week one.
Should a startup do cold outreach or content first?
Outreach to people showing a signal, first. Content takes months to rank and a startup needs conversations this month, both for revenue and to learn how buyers describe the problem. Write the content afterwards, from the questions those conversations raised. Untargeted cold outreach to a bought list is the one thing to skip: low replies and a restricted account.
How many customers does a startup need before the method is validated?
There is no magic number, but a useful test is ten paying customers who did not know you before they bought. Friends and former colleagues prove the product works; strangers prove the message and the channel work. Count signals found, replies and calls weekly, and treat a loop that books calls every week from strangers as the thing to scale.
How long does it take to get customers for a startup this way?
Week one by hand should produce replies and a few calls. In our own launch, with the finding and the first conversations automated, we booked 14 qualified demos in 14 days. How long a call takes to become revenue depends on your buyer: a solo founder decides in a week, a company with a procurement step can take months, so judge the method by calls booked, not by closed deals.
Is it worth paying for a tool before the first ten customers?
Usually not for the sending, often yes for the finding. The sending caps are the same with or without a tool, so the first bottleneck is the hour a day spent re-running searches and reading threads. If that hour is what stops you from doing the plan at all, a tool that watches LinkedIn for you pays for itself in conversations; if you have the hour, run it by hand first and learn the phrases.
