How to get your first customers for a SaaS

Most SaaS founders hit the same wall. The product works, the landing page is live, maybe there was a launch day. And then nothing: a few visits, one or two free signups, no one paying.

That's normal. The first customers of a SaaS almost never come from people finding you. They come from you finding them.

Paul Graham said it plainly in his 2013 essay Do Things that Don't Scale: "The most common unscalable thing founders have to do at the start is to recruit users manually. Nearly all startups have to. You can't wait for users to come to you."

This guide turns that idea into steps you can follow this week. If you don't have a product yet or you sell a service, start with finding your first 10 customers without an audience. This one is for SaaS.

Step 1: pick one buyer who can pay this month

"Small businesses" or "founders" is not a buyer. Write one sentence that names a role, a situation and a pain:

_Example:_ "Office managers at dental clinics with 3 to 10 chairs who still confirm appointments by phone."

Check two things before you go further:

• They have a budget. Someone in that role can pay for a tool without a three-month approval process.
• The pain costs them now. Time lost every week, money lost, a mistake that keeps happening.

A narrow buyer feels like a small market. At the start, that's the point. Paul Graham compares it to "keeping a fire contained at first to get it really hot before adding more logs."

Step 2: find people who are asking right now

SaaS buyers leave public traces when the problem hurts. Look for these, in this order:

1. 1"Alternative to X" posts. Someone unhappy with the tool they pay for is the warmest lead there is: they already have a budget for the problem.
2. 2"How do you handle..." questions in the forums and groups of your niche. The person describes the pain in their own words.
3. 3Workarounds. "I built a spreadsheet to track..." means they want a tool but haven't found the right one.
4. 4Job posts for the problem. A company hiring someone to do the job your tool automates has the pain and a budget.

Places to start: the forums of your niche, Hacker News search sorted by date for "Ask HN" plus your problem, Bluesky search, and Reddit threads. More on what to look for in buying intent signals.

Skip the communities full of other founders selling SaaS. They are your peers, not your buyers, unless founders are your actual target.

Keep only recent posts. A question from last week is a conversation. A question from two years ago is a dead end.

Step 3: start a conversation, not a pitch

Reply to the post or write to the person. Keep the first message short:

• one precise reference to what they wrote;
• one easy question about their situation;
• no link, no price, no demo request.

_Example:_ "Saw your post about confirming appointments by phone. How many calls is that on a normal Monday?"

When they answer, help first: a concrete tip, a template, a way to fix part of the problem today. Mention your tool once, when it fits. The outreach message guide has more examples.

Step 4: set it up for them

This is where most SaaS founders lose their first customers. Someone says "sounds interesting", you send a signup link, and they never finish onboarding.

Do the setup with them instead. At Y Combinator, this has a name. Paul Graham describes how the Collison brothers at Stripe didn't send a link when someone agreed to try the product: they'd say "Right then, give me your laptop" and set them up on the spot.

For a SaaS today, that means:

• a 20-minute call where you share your screen and set up their account with their real data;
• importing their spreadsheet yourself;
• writing their first template, rule or report for them.

Paul Graham did the same at Viaweb: when merchants didn't want to build an online store with the software, the team built it for them. He writes that it "taught us how it would feel to merchants to use our software."

You will learn more in one of these calls than in a month of analytics.

Step 5: charge from the first customer

Free users tell you what they like. Paying customers tell you what the product is worth. The second one is what you need.

Ask for money once they have used the product on their own data. A few ways to do it without pressure:

• a normal monthly price, with a clear refund policy;
• a paid pilot for a fixed period, if the buyer is a company that needs time to decide;
• an early price that you explain honestly, not a fake countdown.

If you are torn between a free trial and a paid pilot, this comparison can help.

If they say no, ask why. "Too expensive", "not now" and "missing one feature" are three different problems with three different fixes.

Step 6: treat the first customers like gold

Your first customers are also your product team. Paul Graham writes that Wufoo "sent each new user a hand-written thank you note" for as long as they could.

You don't need handwritten cards. You need:

• a check-in after the first week: "What did you try to do that didn't work?";
• quick fixes for what blocks them, and a message when it's fixed;
• a direct way to reach you (email or chat with you, not a ticket queue).

For B2B, Paul Graham suggests going further: pick a single user and act "as if they were consultants building something just for that one user." Just don't start charging by the hour. You are a product company being very attentive, not an agency.

Mistakes that slow down the first customers

• Waiting for a launch to work. A launch brings a spike of visits. It rarely brings a steady flow of buyers.
• Building features instead of talking. If no one is paying, the missing piece is usually a conversation, not a feature.
• Selling to everyone. A page that speaks to "teams of all sizes" speaks to no one.
• Free forever too early. A free plan with no limits teaches you how many people like free things.
• Giving up after 20 messages. Count replies, not customers. No replies after 30 messages means change the message or the buyer. Replies but no sales means change the offer.

Your plan for this week

1. 1Write your one-sentence buyer.
2. 2Find 10 recent public posts where that buyer asks for help.
3. 3Reply to 3 of them each day, with a question, no link.
4. 4Offer to set up the product for anyone who is interested.
5. 5Ask for payment once they have used it.

Where FounderPulse fits

Step 2 is the slow part: reading a lot of posts every day to find a few people who are asking right now. FounderPulse does that search for you. It reads recent public posts on forums, Hacker News, Bluesky, Mastodon, Indie Hackers and the open web, checks the date of each post, drops ads and people selling the same thing as you, and tells you in one sentence why each person fits. For each one, it drafts a first message that helps before it sells. You send it yourself.

It doesn't do the calls, the setup or the follow-up. Those are still yours, and they are what turns a conversation into a customer.

You can try it free without signing up: paste your website and see three people to contact today.