How to Find SaaS Customers: A Founder's Playbook

Why early-stage SaaS struggles to find customers

Most SaaS founders don't have a product problem in the first year. They have a distribution problem. The product works, a handful of friendly users like it, and then growth stalls because nobody on the team can answer a simple question: who exactly are the next hundred customers, and where do we find them?

The honest answer is rarely "content marketing" or "paid ads." Both need volume, budget, and time you don't have. What an early-stage company can actually do is pick a narrow segment, build a list of real companies in that segment, and talk to them directly. This article walks through that process end to end: defining your ideal customer profile, finding the places where those buyers already gather, building a workable lead list, and running cold outreach that gets replies instead of spam complaints.

Step 1: Define an ICP you can actually search for

An ideal customer profile (ICP) is only useful if you can turn it into a search query. "SMBs that care about productivity" is not an ICP — you can't look that up anywhere. "Dental clinics with 2–10 chairs in mid-size German cities" is an ICP, because a map, a directory, or a business registry can return that exact list.

Start from evidence, not ambition. Look at your existing users — even if there are only eight of them — and ask three questions:

• Who activated fastest? Which accounts reached your core value moment within days, without hand-holding?
• Who came back? Weekly active accounts tell you more than signups ever will.
• Who paid or asked about pricing? Willingness to pay is the strongest ICP signal there is.

Now describe the overlap in searchable terms: industry or niche, company size, geography, and one or two observable traits (they take bookings, they have a physical location, they hire sales reps). If the description still sounds vague, narrow it until it feels uncomfortably specific. At this stage a market of 2,000 well-defined companies beats a market of 200,000 vague ones, because you can realistically contact the first group this quarter.

Write down the anti-ICP too

Just as important: who should you refuse to chase? List the segments that churned, demanded custom features, or took months to decide. Early teams burn most of their outreach time on leads they should have disqualified on day one. A one-line anti-ICP ("no agencies, no companies without a website, nothing outside the EU") saves entire weeks.

Step 2: Find the watering holes

Before you send a single cold message, map where your buyers already spend attention. Every niche has watering holes — places where operators of that specific business type read, complain, and ask for recommendations:

• Niche communities: Slack and Discord groups, subreddits, Facebook groups built around a profession ("Salon Owners", "Freight Brokers"). Lurk first. The vocabulary people use to describe their problems becomes your outreach copy.
• Review platforms and directories: people comparing your competitors on review sites are in-market right now. Read the negative reviews of incumbent tools — each recurring complaint is a segment with a reason to switch.
• Industry events and associations: member directories and exhibitor lists are pre-qualified lead lists someone else curated for you.
• Local business clusters: for SaaS that serves physical businesses (clinics, gyms, restaurants, workshops), the map itself is the watering hole. Every pin is a prospect with a public phone number.

Watering holes serve two purposes. They give you language — the exact words buyers use, which will double your reply rates later. And they give you lists — names of real companies you can research and contact.

Step 3: Build a lead list you can work through

This is where most founders stall. Manually googling companies, opening each website, and copying emails into a spreadsheet takes three to five minutes per lead. At 200 leads, that's a lost week of founder time — and the data goes stale within months.

A workable lead list has four properties:

1. It matches the ICP, not just the industry keyword. A search for "fitness" returns supplement shops and yoga influencers; you need actual gyms in actual cities.
2. Every row has a reachable channel. A company name without an email, phone, or social profile is not a lead, it's trivia.
3. The channels are verified. Sending WhatsApp messages to landlines or emails to dead inboxes wrecks your deliverability and your morale.
4. It's sized to your capacity. If you can personally handle 25 conversations a day, a 5,000-row list is noise. Build lists of 100–300 and finish them.

You can assemble this by hand from maps, business registries, and web search — or let software do the aggregation. A tool like JustLeadIt, which finds companies by niche and city and collects their public emails, phones, WhatsApp, and social profiles in one search, compresses that lost week into a few minutes — and it verifies which phone numbers actually have WhatsApp before you write to anyone. New accounts get two free searches, enough to test a full ICP hypothesis before spending a cent.

Whatever tooling you choose, export the list to a spreadsheet (XLSX or CSV) and treat it as a working document: one row per company, with columns for channel, first-touch date, and outcome.

Step 4: Cold outreach that works for software

Cold outreach has a bad reputation because most of it is lazy. For SaaS specifically, the bar is higher: your prospects already receive "quick question" emails every day from tools they will never buy. Here is what still works in practice.

Lead with their problem, not your product

The first sentence should prove you know their business. "I noticed your clinic takes bookings by phone only" beats any product pitch. You learned that vocabulary in the watering holes — use it. One observed detail per message is enough; five feels like surveillance.

Make the ask tiny

Don't ask a stranger for a 30-minute demo. Ask a yes/no question they can answer from their phone in ten seconds: "Is double-booking still a headache for you, or did you solve it?" Replies open conversations; calendar links close them.

Pick the channel the buyer actually reads

• Email is the default for software buyers who work at a desk. Keep it under 90 words, plain text, one link at most.
• WhatsApp outperforms email dramatically for owner-operated businesses — salons, clinics, restaurants, trades — across most of Europe, Latin America, the Middle East, and Asia. But only message numbers that verifiably have WhatsApp, and keep volumes low and human.
• LinkedIn and Instagram work as warm-up channels: follow, react, then message. Slower, but almost never resented.

Personalize the first line, template the rest

Hand-writing 200 unique messages doesn't scale; sending 200 identical ones doesn't convert. The working compromise is a strong template with a personalized opener per lead. An AI message generator helps with the first draft, but read every message before it goes out — you're building a reputation in a niche where owners talk to each other.

Follow up exactly twice

Most replies come from follow-ups one and two, sent three and seven days after the first touch. After that, stop. Mark the lead, move on, and revisit the segment next quarter. Persistence past three touches converts almost nothing and poisons the niche for you.

Step 5: Track every conversation or lose it

Past twenty parallel conversations, memory fails. You will re-message people who said no and forget the people who said "ask me next month" — the single most valuable pile in early-stage sales. You don't need a heavyweight CRM yet; you need per-lead tracking of three things: which channel you used, what you said, and what happened. Keep it in your spreadsheet, or in your lead-generation tool if it tracks contact status per lead. Review the "not now" pile monthly; a surprising share of your first paying customers will come from it.

A realistic week-one plan

1. Day 1: Write your ICP and anti-ICP in searchable terms. One sentence each.
2. Day 2: Spend two hours in the watering holes. Collect the exact phrases buyers use to describe the problem you solve.
3. Day 3: Build a list of 100–150 companies matching the ICP, with verified contact channels, exported to a spreadsheet.
4. Days 4–5: Send 20–25 first touches per day on the buyer's preferred channel, each with a personalized first line and a tiny ask.
5. Following week: Send follow-ups, log every outcome, and compare reply rates across segments. Double down on whichever slice of the ICP answers.

Finding SaaS customers at the early stage is not a growth-hacking mystery. It's a definable, repeatable loop: sharpen the ICP, build a verified list, reach out like a human, track everything, and let reply-rate data tell you where your real market is. Run the loop weekly, and your first hundred customers stop being a hope and start being a schedule.