How to Find What to Sell Online With One Prompt
You already have the sellable thing. This is one prompt that names it, prices it, and hands you the message that finds out whether anyone will pay.
The short answer
To find what to sell online, stop looking for a new idea and describe the one you already have. Tell an AI how many years you've worked, your job title, three problems you solved regularly, and who you solved them for. Then constrain it: no website, no funnel, no course, and one idea rather than twenty. Ask it for one problem, a one-sentence positioning statement, exactly who pays, a way to test demand, a message you can send today, and pricing logic rather than a price.
At a glance
- Time
- ≈ 60 seconds
- Cost
- Free plans work
- You need
- Your job history
- Where
- Any AI chat
Here's what I see constantly. People with decades of real expertise, project managers, HR directors, tradespeople, healthcare workers, accountants, teachers, all holding back.
Waiting until the website is perfect. Spending months on social media before making an offer. Building a whole course before checking whether anyone wants it. Everyone's convinced they need the following, the brand, the funnel and the logo before they can do anything at all.
Reading, planning and preparing. Not doing. While already holding the most valuable thing anyone can sell online: proven, paid, real-world expertise.
You've been doing something for 20 or 30 years that people would pay to learn from. And instead of selling that, you're trying to learn something completely new. That's backwards.
Forget passive income. Forget ten thousand a month. Handle the first dollar, because if you can't prove to yourself that someone will pay you $1 for what you already know, the ten thousand was never coming.
Key takeaways
- 01
Your goal right now is proof, not income. You are not trying to build a business this week. You are trying to establish that a stranger will pay for something you already know, because everything else depends on that being true.
- 02
Experience you were paid for is the asset, and you already own it. Twenty years of solving the same problems is the thing people buy. Learning a brand new skill to sell instead is the long way round.
- 03
Constrain the prompt or you get overwhelm. Say no website, no funnel, no course, and one idea rather than a list. A pile of options feels productive and produces nothing.
- 04
Ask for pricing logic, not a price. A number on its own you will talk yourself out of. A number with the reasoning attached survives the first person who hesitates.
- 05
Validate before you build, every time. I ran this on my own marketing system before taking it further, because I needed to know somebody would pay rather than build it and hope.
I know this works because I did it to myself.
I'd been building a marketing system and I needed to know whether anyone would actually pay for it. I couldn't build the whole thing out first and then hope. I had to find out first and validate it.
That step changes how you move forward on everything after it. And for context on whether I'm asking you to do something I find easy: I was taught to type on a typewriter at school. That was the technology they were preparing us for. Now I build AI agents that run my marketing systems. If I can make that jump, the one you're looking at is much smaller.
Prefer to watch
9:40The prompt, and what to do with it
Write down what you already did
Years worked. Job title. Three types of problems you solved regularly in that role. Who you solved them for.
That's the entire input. If you feel stuck listing the problems or the people, upload your resume instead. It already contains all of it, which is rather the point.
Give it the constraints
This is the step people skip, and it's the one that decides whether the output is usable.
Tell it: no website required, nothing built before it's validated, no course, and do not give me a list of ideas. One idea. A list of twenty feels like progress and is actually just overwhelm wearing a hat.
Ask for six specific things
Not "what could I sell". Ask for:
one problem worth solving, a one-sentence positioning statement, a precise description of the specific person who would pay for this, how to test for demand, a validation message you can send today, and pricing logic that explains why that number makes sense.
Read what comes back, and push if it's wrong
If the first idea doesn't fit, ask for the next one. You don't need twenty options. You need one that fits, and you'll know.
Post the validation question
The prompt hands you a message. Post it in a Facebook group, a local group, LinkedIn, or straight to your contacts.
It reads something like: "hey, quick question. If you're dealing with this specific problem, would you pay for a 45 minute call with someone who solved this professionally for 20 years?"
That's it. That's the whole test.
Count the yeses, not the followers
You don't need a hundred people. You need enough to say yes to prove the model works.
Your goal is not income yet. Your goal is proof, and once you have it, everything else opens up.
Two real runs
Twenty years as an HR director
The input was plain. Twenty years as an HR director. Problems solved: handling difficult employees, having hard conversations, resolving conflict. People worked with: small and mid-sized business owners, and managers with no formal HR training.
Nothing invented, nothing polished. Just the job.
What it understood about the buyer
This is the part that impressed me. It didn't say "sell HR consulting".
It said these buyers worry about legal exposure, about the emotional fallout, and about saying the wrong thing. So they will often pay for guidance before the conversation, because the cost of getting it wrong is far higher than the cost of advice.
That's a real insight about when money changes hands, and it's the difference between an offer and a hope.
The positioning, the offer and the price
Positioning: helping small business owners prepare for difficult employee conversations so they're handled professionally, legally, and without unnecessary conflict.
The offer: a 45 minute difficult-conversation prep call. The price: $150 to $300.
And the reasoning, which matters more than the number. Nobody is being asked to learn anything new. It solves a problem that already exists this week for a lot of business owners. Urgent problem, real risk, credible experience. That combination is the fastest route to a paid conversation.
Twenty-two years in construction labour
Completely different trade, same process. No website, no brand.
It came back with helping contractors set up job sites so crews can start working immediately, organised. The offer was a job site efficiency review.
The pricing logic did the arithmetic out loud: ten workers at $40 an hour is $400 in labour. Save half an hour and that's $200 of recovered productivity. Price the review against that.
Where the message goes
For construction, it named the places: local contractor groups, construction Facebook groups, LinkedIn.
Not "promote it". The specific rooms where the specific person already is. That's the difference between a validation test and shouting into your own feed.
Sources
Questions people actually ask
What should I sell online if I have no idea what to sell?+
Sell what you were already paid to do. The prompt works from your job history rather than your interests, because your interests are unproven and your job history is twenty years of evidence that people pay for that skill.
Do I need a website before I can sell anything?+
No, and the prompt is deliberately told to assume you don't have one. A website is a thing you build after somebody has said yes. Building it first is the most common way to spend six months and learn nothing.
What if the first idea it gives me doesn't feel right?+
Ask for the next one. What you should not do is ask for twenty at once. A list is overwhelming and none of them get tested, which leaves you exactly where you started but tireder.
How do I know what to charge?+
Ask for pricing logic rather than a price. The construction example priced a job site review against ten workers at $40 an hour, so half an hour saved is $200 recovered. That reasoning is what you fall back on when someone hesitates.
How do I test whether anyone will actually pay?+
Post one question where your buyer already is. Ask whether they'd pay for a 45 minute call with someone who has solved that exact problem professionally for twenty years. You are looking for yeses, not applause.
How many yeses do I need?+
Not a hundred. Enough to prove the model works, which is a much smaller number than you think and a much bigger deal than you think.
Can I just build a course instead?+
Eventually, and not first. Courses take real time to build, and building one before you know anyone wants it is the same mistake as the website in a different shape. Prove the demand, then decide what to build.
Does this work if my experience is a trade rather than an office job?+
That was the second example, twenty-two years of construction labour, and it produced a clearer offer than the office one. Trades tend to solve expensive problems, which makes the pricing logic easier, not harder.
About the author
Hey, I'm Traci
Digital Marketing Strategist
With over 25 years in this industry I help small business owners and creators turn their online presence into profit.
Keep going
The setup, the templates, and what I build next
Not more reading. The Claude setup guide you can follow tonight, and the templates I actually use. The page opens the moment you sign up.
Get the guideFree · you get it straight away