It's a fair question. ChatGPT can generate prompts. It can even generate prompts about prompts. So why would anyone pay $20 for a pack of text snippets they could theoretically write themselves?
You're not paying for the prompts. You're paying for the editorial labor that came before them.
The hours someone spent asking, "what's the actual recurring task here?" The dozen rejected versions that produced generic output. The constraint that wasn't in version one but appeared in version six. The output format that took three tries before the model would actually follow it.
You can absolutely do that work yourself. Most people don't, because the cost isn't writing one prompt. It's writing a prompt, using it for two weeks, noticing it produces fine-but-not-great results, refining it, and repeating that loop across thirty other recurring tasks in your job. That's a quarter of your professional attention, scattered across small moments when you were trying to be productive anyway.
The hidden cost is iteration, not authoring
When a product manager pastes "help me prioritize this backlog" into ChatGPT, they get a passable answer. They don't know that the same query with a RICE framework specification, a confidence-flag column, and an explicit ranking rule produces a result that's twice as useful and skips the back-and-forth entirely. They don't know because they never wrote the second version. They got the first answer and moved on.
That gap, between "passable first try" and "version six", is what a well-built prompt pack closes. Someone already wrote version six. Tested it across real workflows. Figured out which constraints matter for the specific role.
It's the difference between a recipe and a dish
ChatGPT can give you a recipe for any dish. A recipe is generic. A great dish requires knowing which ingredient to swap if you don't have the original, what temperature your specific oven runs hot at, when to taste and adjust. The recipe is free. The version that consistently produces a great result took someone's evenings for a month.
The prompts in our packs aren't the recipe. They're the version that actually produces the dish. Built by people who use prompts like these in their day jobs, tested across the AI tools customers actually paste them into, refined until the first response is the one you can use without rewriting.
So who should pay for prompts?
If you're a hobbyist, write your own. It's a genuinely good way to learn what good prompts look like, and the friction is part of the lesson.
If you're a working professional whose hour is worth more than $20, buy the version someone else already iterated to v6. The math works out before you even finish reading the first prompt.
That's what we sell. Not the prompts. The version-six prompts. The ones we already broke, fixed, and verified actually work in the real workflow they were built for.