Hi all,
I haven’t been very active here yet, but I wanted to share where I am and get your thoughts.
I used to focus heavily on affiliate marketing and SEO (former WA member, etc.), but over time I got frustrated: despite putting in effort from ~2020-2024, the returns kept decreasing.
I even tried working outside the online business space (as a mechanical engineer) but that didn’t pan out as promised. During that time I took a partial break from online business.
Now I want to come back and work for myself again.
Recently I’ve been exploring prompt engineering (writing and refining prompts for AI tools, perhaps selling prompt templates or offering coaching) instead of more saturated niches (like crypto affiliate).
It seems the learning curve might be more manageable, and there are signs of demand in Czechia.
My plan is:
-
Learn the craft / techniques of prompt engineering.
-
Create content (in Czech) showing how better prompts make better results.
-
Focus on a niche: for example, people applying for jobs, or small-business owners who want help using AI/prompts in their work.
-
Start locally (Czech market), then based on results test and eventually expand to English-speaking environments.
-
Build an audience, test monetization (selling prompt packs, coaching, or services) as I go.
Of course, I have doubts: will this work?
Will I spend several months and see little result?
But I believe with dedication and a clear plan, something can grow.
My question to you all:
Do you think this is a viable business model — especially in Czechia, with possible future expansion to English-language markets?
Any suggestions, pitfalls I should know, or ideas for niches / monetization?
Thanks for reading and any feedback.
Mike
Hi - it's an interesting idea, and something I have thought about, but never followed it up. I can see some potential difficulties, as technology moves so fast, and the market is continually expanding with newer, faster, cheaper AI models.
In fact, you might be surprised to learn, I currently have access to 68 different models, and that's not including the image/video generation platforms! How many people have even heard of the likes of longcat-flash-chat, mai-1-preview, glm-4.5v, step-3 or kimi-k2-0711-preview, for example?
And yet, all of the above give excellent results. So, as you can imagine, getting a good response requires different inputs, they use different numbers of tokens, different API keys. You need to know the rate limits, the token costs, etc.
Of course, all of the above assumes you want to work with companies, in-house or as a freelancer. If you are simply catering to individuals, who have little knowledge of AI, then mastering the basics of the most popular models might suffice.
It's a fascinating and complex subject, and I will be interested to learn how you move forward!
@michalb Michal, I really respect the way you’ve laid out your plan and the fact you’re starting with something local. That part makes sense to me. But I want to be upfront, since you asked if this is viable.
The hard truth is that AI is already very good at writing its own prompts. Some of the most advanced models today can generate and refine instructions better than we can, because they’ve been trained on millions of examples. That doesn’t mean there’s no business here, but it changes where the real value is. Instead of just “better prompts,” I think the bigger opportunity might be helping people apply AI correctly in a specific context - job seekers, business owners, whoever you choose - because they don’t have time to experiment.
It reminds me of a quote I saw floating around online. Something like:
“The advice was to learn coding.
(AI now codes better than humans.)
The advice was then to learn prompt engineering.
(AI now engineers prompts better than humans)
The advice was then to learn plumbing because AI took over all tech jobs!
Then the robots integrated with AI and did plumbing better than plumbers :-)”
Point being, the landscape shifts quickly, as Diane has already mentioned. If you go all-in, make sure you’re not just offering prompts but a way to solve someone’s actual problem. That’s what makes it sustainable.
Curious to see how your Czech market tests go - keep us posted.
You brought up some good points, that Ai can and does write some great prompts of their own.
This includes text content and images.
However, I found that for many, they struggle with finding the right words to convey the image they want.
Because I was one of those folks who struggled, I'm too linear when trying to be creative, visually, I did some research and created a prompt seed starter that provides starter ideas that you can expand on. 150 prompts for 8 different categories.
Without self promoting, the seed prompter is avail on Gumroad.
Between the prompt starters and adding them to the Ai image creator these prompt starters help by providing an idea that you can grow into a usable prompt.
I think it's a great idea to help, Mike, and that's very doable. You've laid out a plan that is flexible when deciding which direction to go.
Rudy
I think Prompt Engineering is a very valuable skillset to have - many people are very limited in how they us AI, or how they input their question/task.
I also think that starting locally (Czech) or at least in a non-English environment is going to help you gain a following a lot more quickly.
Before jumping full-head on into this idea, though, I'd really try to nail down what you think you can sell and/or what you actually want to do.
I see prompts being given away for free all the time - and like every above said, AI moves so quickly, that one day your prompts may work well, and the next they don't.
I think this is definitely an idea that can work, but you just have to figure out how you are going to monetize yourself/your business model before you get too deep into prompt engineering.
@michalb Mike, one angle you might not have considered is how fast attention shifts. A couple of years ago everyone was chasing crypto (many still are of course), then everyone wanted to launch faceless YouTube channels, and now prompt engineering is having its moment. The question I’d ask myself is: how do you make sure you’re not building a business on a passing trend?
From what I’ve seen, the opportunity isn’t so much in the prompts themselves but in packaging. People often don’t want “how to prompt,” they want a shortcut that feels like a finished product. AI agents are likely going to be the next big thing. Lots of potential there.
Since you mentioned Czechia, there might be room to become the “local translator” of these tools and techniques. Most new tech hits the US first, then trickles out slowly. If you can bridge that gap and make it accessible in Czech, you’re not competing head-on with the bigger players.
I’ve been burned before by chasing the shiny thing. It's too easy to get tunnel visioned. If you can figure out a channel where you consistently reach the right people, you’ll be in a much stronger position than relying on the idea of prompt engineering alone.
Keep experimenting ... tools change, but people always pay for convenience.
Leave a reply