Hacker News new | past | comments | ask | show | jobs | submit login
Ask HN: How to Monetize AI App?
4 points by xHashy 10 months ago | hide | past | favorite | 7 comments
We developed a custom LLM model. I have been reaching out to business but I dont see enthusiasm about AI.

How can I monetize my app? It is taking a lot of money to train models so monetization is necessary. ]




What does it do that is different from other LLM models? What problem does it solve? Who has that problem?

Without knowing that, we don't know enough to give any realistic answer. Because businesses don't want "AI". They want solutions to their problems.


A) identify pain that your product solves.

B) identify people experiencing that pain.

C) reach out to those people, showing them how to make the pain go away.

Each step can be hard, or easy.

Firstly, it sounds like you built a product without understanding (yet) your customer. Which can mean you have a "solution looking for a problem". It can be hard to find enough pain in this case because it's harder to find a problem to a solution than a solution to a problem.

Or its easy, your solution generates cash - or some such universal pain.

Secondly finding people with specific pain takes work. Your solution might only appeal to say small town vetenarians. Which means you need to start touring small towns. (Which would be expensive, but you can start I small towns near you.)

Getting people to pay may be hard. They worry about the new pain your app will bring. The existing pain might be slight an thus ignorance.

Generally speaking, it's easier to find problems, and build a solution for them, than it is to build a solution, then look for problems. And ideally you have paying customers signed up before you start constructing the solution.

As creators we want to create first, and find customers later, and yes can work, but you have to be a bit lucky. It's not a sure thing.


Business people won't buy because it is "AI". Don't conflate the opinions of tech bloggers with the pragmatic demands of people running businesses.

A grizzled old business person once said to me: "Show me how it makes me money or saves me money." If you aren't well versed in a specific business domain, you will struggle to demonstrate the benefits of your "solution".

Another problem I have encountered is that large firms (those with the most money to spend) only want to deal with vendors who also are large firms.


> We developed a custom LLM model

Maybe you should ask yourself, why did you do this? Was it to solve your own or someone else’s problem? Or because you thought it would be fun? No shame in the second one, but it’s a lot harder to work backwards from that into a viable product.


Put it behind an API and charge for commerical use or charge after a certain threshold?


Reach out to SaaS companies that could benefit from it.

I am constantly on the look out for models/data sources that would give AIMD an advantage. I am sure others are too.


Create a subscription model with a limited free option so that users can test it. Offer a discount for users if they purchase an annual plan




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: