So I have an idea that might seems crazy at first, but that might be worth it.
I made an algorithm that can generate product ideas using data collected on tens of thousands of startups and products.
I scraped most of ProductHunt and Crunchbase data and fed that into a text generation artificial intelligence(GPT-2). That's more than 200,000 descriptions of products/companies that can be used to generate new ideas.
So now I have a program that can generate virtually unlimited product ideas.
And I found ways of getting free advertising credits and free mass mailing.
I want to automatically find unexplored niches where products are actually needed, and actually build those products whenever I find a potentially profitable niche.
My goal is to generate simple landing pages from those ideas, and to validate them using advertising or cold mailing.
Most generated landing pages are completely worthless, but since I have virtually unlimited advertising, I want to filter them using the wisdom of the crowds.(I would also pre-filter them manually, but what matters most imo is the click through rate and the conversation rate).
How would you proceed in my position? While minimizing costs?
Do you think that I need custom domain? Or can I launch an ad campaign using a subdomain like XXXX(dot)landen(dot)co or producthunt(dot)com/upcoming/MyCompany?
If that's a problem, could I use free domain names without destroying my conversion rate?(for example, domains like .cf .ml .ga. .gq, free african domain names, I kinda want to use dot ml domains for AI related products so I don't seem like a cheap ass)
Do you think that 500 characters is enough content for a simple landing page?Or is that not enough content?
Also, if this was made public would you be interested?
For example a few dollars a month for unlimited AI generated landing pages?(everyday you'd get mailed, IDK 10, 20, 50 AI-generated landing pages on themes that you care about, you could edit them and launch them yourself)
Or a few dozens bucks/months for landing pages that already interest a lot of people?
(everyday you could get mailed 10 landing pages that were already tested, in industries that you care about(for example, education, art,gaming....)), you'd get 10, 500 characters landing pages, with a click through rate of +10% and +5% conversation rate on google ads or something similar.
IDK if that's a good idea, and if I'm completely clear about what I'm doing, but I think it has potential
I would use it differently:
Then you have interesting data and you can play with
etc...
As @Dedupely mentioned, I would be interested in some examples what you got.
Most people have a lot of ideas. I would start publishing lists of ideas on various platforms and see if they generate any interest.
I can see how your got your moniker MakingThingsWork. Please do share a few of the product ideas generated by this algorithm, so we can confirm it's not a monkeys-typing-Shakespeare affair.
You have a business model on your hands with that free unlimited advertising scheme - focus on offering that, then evolve into the ML-powered unicorn printer
So ambitious, admirable!
Thoughts
Let me know when you have something to try out, cheers!
Free ads and free mailing can be a product itself --- bingo.
Sounds like a really cool idea.. Can I give it a try? Most specifically, just generate a product idea?
This comment was deleted 3 years ago