How to do founder sales with no SDR

You do not need a sales hire to get your first buyers. You need a product someone will pay for, and you need to ask those people yourself.

In 2026 you can build an app fast. Pieter Levels said the hard part in June: everyone can build apps, and almost nobody has an audience, ad money, or a free way to get attention. In August he shortened it: more apps than ever, less sales than ever. Shipping is cheap. Another feature will not sell the product for you. An SDR will not invent the buyers either.

Paul Graham wrote this in 2013, and it still holds. You cannot wait for users to come to you. You have to go out and get them. Founders skip this for two reasons he names: they would rather write code than get rejected, and the first numbers feel too small to matter. Neither reason gets you a customer. At least one founder spends a lot of time on sales and marketing. You cannot skip that by hiring a salesperson first.

You can't wait for users to come to you. You have to go out and get them.

Graham is blunt about the hire. You have to do sales yourself at first. Later you can hire a real salesperson to replace you. An SDR is not that hire. Indiecity is for one person or a team under twenty. Name what an SDR actually does, then run the week yourself.

An SDR is not a sales department

An SDR books meetings. That is the job. They work a list, send cold emails, and try to put someone on your calendar. They do not write your pitch. They do not set your price. They do not know who pays.

Those three things are the sale. If you have never closed, you do not have them yet. You learn them on calls you run. A meeting with a pitch you cannot say, at a price you cannot defend, to a person who cannot buy, is a calendar entry. It is not revenue.

Jason Lemkin answered whether a founder should handle sales at the start. The answer is yes, 95% of the time. The founder should close at least the first 10 paying customers (or 20, or whatever that first set is). It is fine if you are terrible at it. What matters is that you still get those paying customers closed. Then hire the first two reps, not one. Hiring before you closed 10 is, in his words, super risky. You cannot tell who is good at the job if you have never done it.

In September 2016 a commenter, Casey Schorr, asked the hire you are weighing: take an SDR first, and pass meetings to the founder, because an SDR is cheaper than a closer. Lemkin’s article still says the founder closes first. Hiring an SDR at zero customers copies how a funded team is already staffed. Those teams already closed. They already know the pitch, the price, and who pays. If you sell, you need names, not a title.

Write thirty names before you hire anyone

Patrick McKenzie wrote it for Stripe Atlas: if you build it, they will do absolutely nothing. You recruit first customers one at a time, with active sales. You start with a spreadsheet. His columns are company, person, and email. The sheet starts blank. You put names on it. You do not need many to get started. A few dozen will keep you busy.

Write thirty named people before you hire anyone. Use these columns:

• Name
• Company
• Role
• How you know them, or the signal that they might buy
• Buyer or bridge
• Learn or sell
• Next step
• Date

Name, company, and role stop you from writing “agencies” or “founders.” How you know them (or the signal) tells you why this person, not a job title. Buyer or bridge is the split that matters: this person can pay, or this person can introduce you to someone who can. Learn or sell is the other split: you still need their week, or you already have a price and an ask. Next step and date turn the row into work. A sheet with no next step is a list you will not use.

Warm names go first. McKenzie is clear: you close more from warm introductions than from the same number of cold pitches. Start with colleagues, past clients, and people in your network. That is how to get SaaS customers from people you already know. If they will not buy, they can still introduce you. After the warm names, write the industries and roles you expect to use the product, then search until you have real people.

If you cannot write thirty, you do not know the buyer yet. That is the test. An SDR cannot invent a buyer you cannot name. Fix the sentence in how to describe your SaaS in one sentence, then come back to the sheet. Not every name is a sales call.

Split interviews from sales calls

Mark a name as learn when you cannot say the problem in their words yet, or you have never heard what they already tried. Mark a name as sell when the pain is clear, you have a price, and the next step is an ask. Mix the two and people stay polite. They do not pay.

A learning call is about their week and what they already tried. Talk about their life first. Do not open with the product. Rob Fitzpatrick’s Mom Test is simple: opening with the pitch changes a learning conversation into a sales conversation. Later you come back and try to get them to use it or pay. The first minutes, before you mention the product, are when you learn the most.

A sales call has a price and an ask. You say what you do, what it costs, and whether they want to start this week. If you still need their words, run customer interviews when you have zero users with the names marked learn. If they praise the idea and you walk away happy, you collected a compliment. Use how to tell if people will pay when “that sounds great” starts feeling like a yes.

Write one next step on each row. Then put the work on a calendar so it happens.

The week: what, when, and where

Here is a week one person can keep. McKenzie’s Stripe Atlas guide says to set dedicated time to get names on the sheet. Put those hours on your real calendar. Do the same work in the same place each day so you do not invent a new system on Tuesday.

Monday: the list

What: finish or rewrite the thirty. Where: one spreadsheet, nowhere else. When: one block in the morning, before you open the editor. Each row needs name, company, role, signal, buyer or bridge, learn or sell, and the next step. The signal is the real reason they are on the list: a complaint they made, a job that owns the budget, a project you shared. If you cannot write the signal, you do not have a reason to write them this week. Cut them or move them down.

Tuesday and Wednesday: the notes

What: five to ten human notes a day, only as many as you can write as a person. Where: your real inbox, or a DM you already use with that person. Not a sequencer. Not a domain you bought just to send cold mail. When: the same two hours each afternoon. One note, one person. In 2026 the same AI opener is already in their inbox. A line that only fits them still gets opened.

Thursday: the calls

What: 15 minutes. Where: your calendar. When: one afternoon block, not scattered leftovers. Send the booking link after they say yes, not in the first mail. Learning calls stay about their week and what they already tried. Sales calls end with a price and an ask. Keep Thursday for calls. If you keep writing notes all week, you never take the calls.

Friday: the count

What: follow up by hand, mark the sheet, count sent, replied, booked, and paid. Where: the same spreadsheet, then one room where buyers already talk if the list is thin. When: Friday afternoon, before you go back to the editor. Sent proves you worked. Paid proves a sale. Do not buy a list to fill Friday.

1. Monday, spreadsheet: rewrite thirty rows until every name has a signal and a next step.
2. Tuesday, real inbox: send notes to buyers you know.
3. Wednesday, same inbox: send intro asks to bridges, then the first human cold notes if the warm list is thin.
4. Thursday, calendar: 15-minute calls. Learn or sell. No mixed pitch.
5. Friday, spreadsheet plus one buyer room: follow up, count four numbers, add names you earned.

That is the week. Start the notes with people you already know.

Start warm

The first notes go to people you know, or to people who can intro you. Colleagues, past clients, anyone you can name without buying a list. If they will not buy, ask them for an intro.

Bridges are not customers. A past coworker who will never pay is still useful if they work with the buyer. Ask them for one name, not a purchase. Do not ask them to invent a pitch. If a warm name is marked learn, ask about their week. Do not pitch. If you cannot say the offer in one line, write it in One sentence before you send. If you only need one yes this week, stay on how to get your first paying SaaS customer until money moves.

Hey [name]. Last time we spoke you were dealing with [messy thing in their words]. I built a small tool that does [outcome] for [who]. It is [price]. If that is still a pain, I can show you in 15 minutes this week. If it is not, who should I talk to instead?

Hey [name]. I am looking for [job title] who still deals with [pain in their words]. I charge [price] to [outcome]. Do you know one person I should write to? A name is enough. If you want, I can send three lines you can forward.

Send the buyer notes and the intro asks first. Send them in the channel you already use with that person. When the warm list is thin, write one human cold note. Do not buy an SDR tool.

The note is four sentences, not a sequence

When the warm list is thin, write one cold note to one person. If they already answer you, skip the cold note and use First ask. A stranger gets four sentences and one ask.

McKenzie’s Stripe Atlas guide still says start with a three to four sentence cold email whose only job is a short call. One signal. One outcome. One Thursday ask. His last line is the one to use: “Do you have 15 minutes free this Thursday?” Leave the calendar link out of this mail. Send it after they say yes. Do not add a backup day in the first note. If Thursday is wrong, they will say so.

• Signal: a public fact that proves they have the mess now.
• Outcome: the one thing you will show, not a product tour.
• Ask: 15 minutes on a named day.
• After they say yes: the booking link. Not before.

If they stay quiet, write the follow-up by hand, once or twice a week. Atlas is blunt: treat silence as silence, not as a no. You can do twenty follow-ups in ten minutes if you are organized. That is the follow-up job, not a sequencer. For the first line, use how to write a cold email that books a SaaS demo.

Hey [name]. I saw [specific signal: the post, the hire, the tool, or the complaint] and it sounds like [pain in their words] is still on your plate. I built a small tool that does [one outcome] for [who]. Do you have 15 minutes free this Thursday?

When they say yes, send the link and stop pitching in mail. The call is a different job.

Fifteen minutes, then the price

They talk first. Ask what they do today and what it costs them. Then show one outcome. If you cannot name the one thing you will show before the call starts, you are not ready.

Before you hang up, say the price and stop talking. McKenzie wrote in 2006 that solo makers undervalue their software. Price against what the buyer already spends or saves, not against how ugly the build felt. If you have no rival to copy, price against the time and tools they already burn. The price from value tool will give you a draft number. Edit it. Then say it.

Ask, point-blank, if they can independently decide to buy. Atlas tells you to ask that. If they will not give a straight yes, treat it as not now and go to the next name. Do not hide the ask behind a waitlist. A waitlist is not a sale. Use waitlist vs charging on day one if you are still hiding the number. Do not open a free trial to dodge the number. For the first few, Atlas prefers helping them start over a trial, because a trial is an easy way to never begin.

If this takes [pain] off your week, it is [price]. I can start you today. Can you decide that yourself?

Fine. Let’s pick a time this week to finish it: [day] [time] or [day] [time]. If you decide no, one line is enough.

What do you already spend on this mess: time, a tool, a person? I priced this against that, not against how long I took to build it. I can start you on [shorter start] at [price]. I will not make it free.

A polite look is not a buyer. If they only wanted education, write that down and use how to tell if people will pay, not just say they like it. The week is not over when the call ends.

Take the money the same day

Send the payment link, the invoice, or the transfer before you close the laptop. Paul Graham wrote that when anyone agreed to try Stripe, the Collison brothers said give me your laptop and set them up on the spot. The same essay still holds: you have to go out and get them. You do not send a beta link and wait.

For the first few, help them start. Import the file. Push the buttons. Do not send a trial and hope they come back. After they pay, ask for one intro. That is how the first paying SaaS customer becomes two.

Thank you. I will get you live today. Who else has this same mess? A name and a short intro is the most useful thing you can do for me right now.

You only hire after you can write those steps down. McKenzie wrote in 2014 that for years he had no sales process. Quotes hit the inbox. He emailed a price. He later wrote that he should have asked that buyer onto a call. He dropped follow-up. He only built a process someone else could run after he could describe the steps himself. That is why you do not hire an SDR yet. You cannot hire someone to run a week you have not written down.

Four numbers, not a CRM you never open

Count the week with four numbers: sent, replied, booked, paid. A sheet is enough.

• Sent: notes you wrote and sent
• Replied: they wrote back
• Booked: a time on the calendar
• Paid: money moved

Replies without money are curiosity. Someone said it was interesting. That is not paid. If ten or twenty people who fit stay silent, change the buyer or the offer. Do not hire an SDR to send more of the same. Silence is an answer about the list. It is not a reason to hire.

You still do not hire an SDR

You hire a closer later, if ever. Hire after you can walk someone through the call. Not after one quiet week. Lemkin’s first-10 rule is the bar. In March 2024 he added that the biggest mistake when you hire a first VP of Sales is stepping out of sales. Every other first VP needs you doing at least half of what you were doing.

Rob Walling’s Core Four (November 2025) is sales, marketing, product, and development. People ask if they can just hire the missing one. He said no: it will not be faster, and it is much more likely to fail. Get well enough at all four and you do not need a cofounder. If you refuse to learn sales, get a founder-level sales person, or pick an idea that does not need sales calls. That is not an SDR this month.

Arvid Kahl wrote in May 2025 that after about a year, near break-even, he was already running demo and onboarding calls for Podscan. The first hire he wanted was not an engineer. It was a part-time or commission seller for non-technical buyers, because he could not pay a full-time hire. He sold himself first. That is a later hire, after the founder can describe the call.

Paul Graham wrote in 2015, about companies that raise, that hiring too fast is by far the biggest killer. Fix a moderately appealing product by doing things that do not scale, not by adding people. When Canny thought about outbound sales as a first hire, other founders asked which channels already worked. Sales was not one of them. That story is in when to hire your first employee in a bootstrapped SaaS. Until you can teach the call, run the week. Do not hire.

What to stop doing

• Hiring an SDR before you have closed, or before you can walk someone through the call.
• Pasting the same generated opener to twenty people.
• Putting a calendar link in the first mail.
• Mixing a learning call with a pitch.
• Buying a list to fill Friday.
• Opening a free trial so you can skip saying the price.
• Waiting for a CRM before you send the first five notes.

If the week was quiet, change the buyer or the offer. Do not hire someone to send more of the same note.

Run the same week twice

Repeat the week. Same notes. Same four numbers on Friday. Then use how to get your first 10 paying customers without ads for the longer loop: rewrite the page with their words, send the next names, and ask every paid customer for one intro. After ten, the first 100 is the same week with more names, not a sales team.

When someone pays, write it down. You can join Indiecity while you run the week. You will have people to talk to. It does not send the notes. If the product has a name and a page, put it on the map.