The first salesperson you hire should be you, for longer than you want

The wish is universal and completely understandable: hire someone who is good at selling, so the founder can go back to building the thing. It is also the most reliably expensive early hire in software, and the reason is not that salespeople are bad at their jobs.

I cannot imagine anyone but the founder doing the sales. I don't know of a single instance where that has worked.

Rob Walling

The ladder

One practitioner's thresholds for what may leave the founder, and when:

Revenue	What can be delegated
Below $300k	Nothing sales-related
$300k – $1M	Appointment setting only
$1M – $3M	Salespeople
$3M+	A VP of Sales

Below $300,000 you are not running a sales process, you are discovering one. Every call is teaching you which objection is real, which segment converts, what people call the problem, and what has to be true before anyone signs. A hired rep can execute a process. They cannot discover one, because discovering one requires changing the product and the positioning in response to what you hear, and that is not a thing you can delegate.

And the universal gate, before any hire at any level: run the full process yourself at least ten times and document it end to end. Not ten calls — ten complete cycles, from first contact to signature or death, written down. If you cannot hand a new person that document, you are not hiring a salesperson. You are hiring someone to invent your sales process while being paid commission on it.

Hire two or none

When you do hire, there is a counting argument that saves a year:

If you can't afford two, you're not ready for one.

Damian Thompson

One rep gives you an argument. Two give you a diagnosis:

Outcome	Diagnosis
Both fail	The market, the product or the process — not the person
Both succeed	It works. Scale it
One of each	A personnel issue, and now you know

With a single rep, every disappointing quarter produces the same unresolvable debate: is it them, or is it us? You cannot answer it, so you either fire someone who was right or keep someone who was wrong, and both cost a year.

One related rule: never hire commission-only. It reads as risk-free and it is not — it signals to the candidate that you do not trust your own process enough to fund it, which selects for exactly the people who will not stay.

Learning first, scale last

Which channel to use early is not a scale question either. The order that produces the most learning, from most context to least:

1. In person. Highest bandwidth, lowest volume. You see the face when the price lands.
2. Cold calling. Real-time objections, still conversational.
3. Cold email. Scale, and almost no signal per attempt.

In the early days I wouldn't worry about what does scale or not. I would focus more on how I can learn as much as humanly possible about my customer.

Steli Efti

This is the opposite of how most founders sequence it, because email is the one you can do without talking to anybody. It is also the one that teaches you the least per attempt — a non-reply carries no information at all. Fifty conversations tell you more than five thousand sends, and they take less calendar time than you think.

The activity model, with numbers

Founder-led sales sounds vague until someone counts it. One worked example: 104 conversations produced 24 sales calls, which produced $10,000–20,000 a month in recurring revenue — a 23% conversation-to-call rate.

That is a usable model. It says roughly four conversations per booked call, and it sets an honest expectation for the top of the funnel: if you want two calls a week, you need about eight conversations a week, every week. It also reframes what the work is. The unit is not the email sent, it is the conversation had — and 104 of them is a quarter of deliberate effort, not a heroic year.

Pair it with the win rate you should expect once the calls happen: you will lose 75–85% of first meetings, and 25% is a very good win rate. Nothing in those numbers is a sign that something is broken.

If you hate selling

The most common private objection is not about strategy. It is that the founder does not want to do this and does not believe they are any good at it.

The answer is not a personality transplant. It is a script:

I'm the founder. I'm really not a salesperson, but I know you wanted a demo — talk to me about your setup today.

A founder's opening line

Two sentences. It sets expectations, removes the pressure to perform, and hands the conversation to the buyer, which is where it should have been. Buyers on an early call are not looking for polish. They are looking for someone who knows the domain and will answer a direct question directly — and a founder outperforms a professional seller on both.

What actually makes this easier is preparation rather than charisma: the ten objections written down with short answers, a handful of specific examples you can recall rather than invent, and the price said out loud on the first call. Those three remove most of what makes selling uncomfortable, and none of them requires being an extrovert.

Five things

1. Do not hire until you have run ten full cycles and written the process down.
2. When you hire, hire two — or wait until you can.
3. Sequence for learning: in person, then calls, then email.
4. Count conversations, not sends. Four conversations per booked call is a workable planning assumption until you have your own number.
5. Write the opening line you will use when you are dreading the call, and use it. The script is the fix.