Cold Outreach for Startups:
How to Land Your First 10 Customers

Here's something nobody tells you when you're building: your first 10 customers won't come from a landing page. They won't come from SEO, or a Product Hunt launch, or a clever tweet. They'll come from cold outreach — from you, personally, reaching out to the right strangers at the right time with the right message.

That sounds obvious. But most founders either skip it entirely ("I don't want to be spammy") or do it so badly it doesn't work ("Here's our product — let me know if you're interested!"). Neither gets customers.

Here's the exact framework I use. It works when you have zero social proof, zero brand recognition, and zero marketing budget.

Why Cold Outreach Works When Nothing Else Does

At zero traction, every other channel has a catch. Ads need money and a proven conversion funnel. SEO takes six months before it moves. Content marketing compounds over years. Cold outreach is the only channel that works on day one — because it doesn't require anyone to come to you. You go to them.

The math is simple: if you can find 50 people who fit your ideal customer profile and send them a message that resonates, a 10% reply rate gives you 5 conversations. Two or three of those convert to customers. Do that twice and you have your first 10.

Step 1: Get Your ICP Right Before Touching a List

ICP stands for Ideal Customer Profile. Most founders define it too broadly: "SMBs in the US" or "SaaS companies." That's not a customer profile — it's a market segment.

A real ICP has three layers:

1. Firmographics: Company size, industry, funding stage, tech stack. The more specific the better. "Bootstrapped B2B SaaS founders with 1-5 employees" is infinitely more actionable than "small businesses."
2. Role: Who feels the pain you're solving? Is it the founder, the head of marketing, the ops lead? Title matters less than whether they own the problem.
3. Trigger: What's happening in their world that makes them ready to buy? Recent funding, new hire, product launch, regulatory change — triggers are what separate "vaguely interested" from "ready to pay."

The fastest way to build this: look at your first five beta users, paying customers, or enthusiastic free users. What do they have in common? That's your ICP.

Step 2: Build a Tight Prospect List

Quality beats volume by a wide margin. A list of 50 people who perfectly fit your ICP will outperform a list of 500 people who sort of fit it. Spam filters punish volume anyway — personalized outreach to a small list performs better on deliverability and reply rate.

Where to find prospects:

• LinkedIn: Sales Navigator is worth $100/month at this stage. Filter by title, company size, industry, and growth signals.
• IndieHackers / Twitter / X: If you're targeting founders and indie hackers, they're broadcasting their problems publicly. Search for pain-point language related to your product.
• ProductHunt: Companies that launched recently are actively in growth mode and open to tools that can help.
• Job boards: A company hiring for a specific role signals what problems they're trying to solve. Someone posting for a "head of outreach" is probably doing outreach badly right now.

For each prospect, note: their name, company, role, and one specific observation about their situation. This observation becomes your email opener.

Step 3: Write the Four-Email Sequence

Cold outreach at the startup stage isn't a single email — it's a sequence. Most responses come from follow-up 2 or 3, not the first email. A four-email sequence over three weeks is the sweet spot: persistent enough to get attention, not annoying enough to earn a block.

Email 1: The Opener (Days 1-2)

This is your hardest email to write because you have zero social proof. No logos, no case studies, no "we helped 50 companies." You're asking a stranger to give you time based purely on the relevance of your observation and the plausibility of your offer.

The formula that works:

Notice what's missing: company description, feature list, pricing, social proof, "we help companies like yours." None of that. At this stage, a short targeted email beats a complete sales pitch every time.

Email 2: The Value Drop (Day 5-7)

Don't follow up with "just checking in." That's noise. Follow-up email 2 should add something: a piece of data, a specific insight, or one concrete thing you can offer. Frame it as the reason you're following up, not a reminder that they didn't respond.

Email 3: Social Proof or Case Study (Day 10-12)

By email 3, you should have something real to share. A beta user result, a specific metric, a brief story. Even one data point — "one of our beta users cut their sequence setup time from 4 hours to 20 minutes" — adds credibility you didn't have in email 1.

If you genuinely have nothing yet, be honest about it. "We're still early — but here's what we're seeing in beta" works better than a vague claim.

Email 4: The Break-Up (Day 16-18)

The break-up email has a counterintuitive effect: it gets more responses than any other email in the sequence. Because it's the last one, it removes pressure. People who were mildly interested but didn't prioritize you will respond when they think it's their last chance.

Keep it short: "I'll stop following up after this. If the timing's off, totally understand — I'll reach out in a few months." That's it.

Step 4: Timing and Volume

Cold email timing matters more than most founders realize. The best windows are Tuesday through Thursday, 7-9am local time for the recipient. Monday mornings are cluttered. Friday afternoons get punted to Monday, where they die.

Volume: start with 20-30 personalized emails per week. This is low enough to personalize meaningfully and high enough to generate data. At 20 emails/week, you'll have 80 conversations in a month — more than enough to find the message that converts.

Don't scale volume until your reply-to-meeting rate is above 20%. If you're scaling a broken sequence, you're just burning a list faster.

Step 5: What to Say When They Reply

Most founders lose deals here. Someone replies with "interesting, tell me more" and the founder sends a six-paragraph product overview. Wrong.

Reply with one question that qualifies them: "What's your current process for [thing you solve]?" That's it. You want them talking, not reading. The goal of the reply is to schedule a call — that's the only conversion event that matters at this stage.

If they say no: "Totally fair — is there a better time to check in, or is this just not a fit right now?" Half the time, "not right now" means three months from now with a different budget situation.

How I Built the Sequences (And What I Use Now)

When I started doing cold outreach for my own projects, I spent two hours per sequence — writing four emails, timing them, figuring out which follow-up angle to use. The math didn't work at scale: too much time per prospect, no consistency across sequences, and no way to know which emails were actually working.

That's why I built 99 Agents. You put in your company and target customer, and it generates a timed four-email sequence — opener, value drop, case study, break-up — calibrated for your ICP. It also runs a deliverability health check on every email before you send, so you're not landing in spam before you even start.

It won't replace the personalization work in Step 2. You still need to understand your ICP and find the right prospects. But it short-circuits the part that kills most founders: staring at a blank draft for two hours every time they want to run a new sequence.

The Only Metric That Matters at This Stage

Forget open rates. Track reply-to-meeting booked rate. If you're getting replies but no meetings, your emails are interesting but not compelling. If you're getting meetings but no closes, the problem is elsewhere. But if you're not getting meetings, nothing else matters.

At 10 emails per week with a 10% reply rate, you get 1 reply per week. At 40 emails per week, that's 4 replies. Four calls per week is plenty to close 2-3 customers per month. In three months, you have your first 10.

The founders who get first customers aren't better at product. They're just willing to send the emails.