Vibe-coded it. Now sell it: the paid-pilot playbook for solo AI founders
Building the product took a weekend. Selling it takes a different skill, and most solo AI founders discover this on the first call, when a buyer asks a question the demo cannot answer. Early sales is not charisma. It is a small number of decisions made before the call and a small number of lines said during it. This is the playbook: who to call, what to charge, how to structure the pilot so it converts, and what to say when they push back.
Sell a paid pilot, not a free trial: 4–8 weeks, a fixed fee (20–35% of the annual price is common), one outcome the buyer chose, a named owner on their side, and a written conversion step: “if the outcome is met, we move to the annual plan at $X.” Then demo only the outcome, name the price, and stop talking.
Why a paid pilot and not a free trial
A free trial asks the buyer to do the work of evaluating your product for nothing, so they don’t. A paid pilot asks them to commit a small amount of money and a named person, which is what makes the evaluation happen. The fee is not revenue; it is the signal that someone on their side will show up to the weekly call. Founders worry that charging will lose the deal. In practice it loses the deals that were never going to convert and keeps the ones that were.
Price the pilot at a fraction of the annual price you intend to charge; 20–35% is common. Say both numbers in one breath — “the pilot is $2,400 for six weeks; the annual plan it converts into is $9,600” — because naming the destination price now is what makes the conversion conversation short later.
The four terms every pilot needs
A pilot needs four terms in writing: one outcome, a named owner, a fixed length and fee, and the conversion step.
| Term | Write it as | Why it matters |
|---|---|---|
| One outcome, chosen by the buyer | “Success means: renewal calls placed before the competing quote lands for 80% of clients due in the period.” | A pilot with three goals is judged on the one that failed. |
| A named owner on their side | “Dana owns the pilot and joins the Tuesday 20-minute check-in.” | Without a name, nobody uses it and the pilot fails quietly. |
| A fixed length and fee | “Six weeks, $2,400, invoiced at the start.” | Open-ended pilots do not end; they fade. |
| The conversion step, in writing | “If the outcome is met, we move to the annual plan at $9,600 on week seven.” | The pilot sale and the annual sale become one decision. |
Put all four on one page and send it after the call. A buyer who will not agree to a named owner has told you something useful before you have spent a week onboarding them.
The ten calls: who, and what to say
Find ten people who have the problem your demo solves — not ten people who might find it interesting. The difference is visible in the first sentence of the call: “You said in your post that renewals are where you lose clients” lands; “I built an AI tool for agencies” does not. Get the first ten from the places the problem is already being discussed: the forum thread, the comment under the competitor’s launch, the person who complained on a podcast.
On the call, three moves. Open with their problem in their words, then ask what happens today. Demo the outcome, not the product — the moment the tool does the thing, and nothing before it; a tour of the settings is where founder demos die. Ask for the pilot with the four terms, and be quiet after you name the price. The silence is the buyer deciding; filling it with a discount ends the decision.
Before the first call, write down what you are selling: the name, who it is for, the pilot price, the annual price, the floor you will accept, the two proof points you have. Keep it in front of you. Founders change the price mid-call because they never wrote it down.
Put your product on file once; NextSay’s Copilot surfaces the next move when price or “can’t ChatGPT do this” comes up, and the Debrief keeps the commitments made on the call, verbatim.
The objections you will hear, and the line for each
Five objections come up on almost every founder call. Each line below buys the buyer’s real concern; the concern is what you then answer.
“Can’t ChatGPT do this?”
“For part of it, yes. Which part are you picturing — and what happens to the rest?” Then name the rest. The data it would need, the integration, the accuracy on their cases, who is accountable when it is wrong, the record it leaves behind. Do not discount in response to this objection; a lower price confirms the comparison.
“How do I know it works on our data?”
“You don’t yet — that is what the pilot is for. Which twenty cases would convince you?” The pilot outcome should be defined on their cases, not your benchmark.
“It’s just you — what if you disappear?”
“It is just me for now, which is why the pilot is short and the fee is small. Here is what you keep if I vanish: [data export, the record, the integration].” Do not promise a team you do not have.
“Let’s start free and see.”
“I’d rather we both had a reason to show up. The fee is small; the named owner is the part that makes it work. Who would that be?” If the answer is nobody, the pilot would have failed anyway.
“We need security review / procurement first.”
“Understood. What is the shortest path — a data-handling one-pager, or a pilot under the threshold that skips it?” Many pilots are priced under the procurement threshold on purpose.
Rehearse the five aloud before the first call; the one you have not practised is the one you will hear. When the line surfaces the real concern, answer it specifically or say you cannot.
Frequently asked questions
How much should a solo founder charge for a pilot?
A fraction of the annual price you intend to charge — 20–35% is common — for 4–8 weeks, invoiced at the start, with the annual price named in the same sentence so the conversion conversation is short later. The fee is a commitment signal, not revenue.
What should a paid pilot agreement include?
Four terms on one page: one outcome chosen by the buyer, a named owner on their side who joins a weekly check-in, a fixed length and fee, and the conversion step. Write the conversion step out: “if the outcome is met, we move to the annual plan at $X on week seven”.
How do you answer “can’t ChatGPT do this?” as an AI founder?
Agree that it can do part of it, then ask which part they picture and what happens to the rest. Name the rest: the data it needs, the integration, accuracy on their cases, accountability when it is wrong, the record it leaves. Never discount in response; a lower price confirms the comparison.
Should an AI startup offer free trials or paid pilots?
Paid pilots. A free trial asks the buyer to evaluate your product for nothing, so they don’t; a small fee and a named owner are what make the evaluation happen. Charging loses the deals that would not have converted and keeps the ones that would.
Price the pilot, name the owner, write the conversion step.
Put your product on file once and let NextSay cue the live objection and keep the commitments from the call.