How do GPT Builders make money?

So with the GPT Store launching the revenue sharing is not ready yet. It will take a couple month and initially it will be for USA only.
How then do developer make money? It seems like GPTs need to authenticate users and then charge for subscription…

2 Likes

You need a Plus account to use GPTs, they are already authenticated.

I don’t understand how developers expect to make money off GPTs. I’m looking at GPTs built by sites that sell subscription but they don’t authentic me as the user (even though they could). What is everyone hoping for?

With 180 million ChatGPT users and 250k Plus subscribers [source], the market seems vast. But, let’s crunch the numbers. Assuming a 10% revenue share, the theoretical upper limit for all creators if they reach every Plus user is $6 million annually. A Plausible Scenario of a Successful GPT:

  • 1% of Plus users engage with one of your GPTs.
  • These users are equally engaged with 5 GPTs from different creators.

OpenAI earnings per Plus user: $20 / mo
Annualized Plus revenue: $20 * 250,000 * 12 = $60 mill
Revenue share for all creators: 10% of $60 mill = $6 mill
Your final share as one of five creators used by 1% of Plus users:
$6 million / (100 * 5) = $12,000 Annually

The 10% estimation is based on my opinion, that given the resources spent are OpenAI’s and not the user’s (like an appstore), the share will be quite lower.

I wrote a bit more about what creators can get to make some money on this model, and it’s mostly going to be companies that just use GPTs for engagement with their services.

1 Like

Nerdnav, an affiliate marketing media site is your source for the user numbers… that seems less than optimal. 250k seems very small.

1 Like

There is no public source. Make do with what we have.

Still, it’s based on the revenue data available. Which could be read as being from Mobile+Plus, but still. I don’t think we’ll jump more than an order of magnitude, which still leaves top creators making very very little in comparison to other 2-sided marketplaces.

Granted, (and I appreciate your pushback) if the number is 10% of users are plus (which would already make them the most valuable private company), the earnings for the “Big winners” could be 120k a year… which is a third of what you can get with a normal job programming in the Bay area. And this is by building a tool that can be replicated pretty easily. No moat and very little reward.

1 Like

I am pretty sure it’s a bit under 10 million.

This forum has half a million users and they are a very small fraction of people that have made use of the API, an even more niche and pay to use group, so… I think that number is not a bit off I think it’s way off.

2 Likes

I would argue there is no direct correlation between those that have logged in at least one to the only active support they can find for their issue (be it API, prompting, announcements…) and those that pay 20$/mo consistently for Plus.

But then again, just some estimate based on available information. I would love to get closer to a real number. How would you estimate the total addressable market of the GPT Store for creators?

I’d say its a million+ as a minimum.

(more words)

There are many factors you can use, GPT conversations, forum members, OAI Staff members. One of the most popular GPTs has 400,000+ conversations. Most of the time people have one conversation with a GPT get bored then never use it again.

1 Like

Using your preferred from any or all of those, what would be your current estimate of paying Plus, Teams & Enterprise users?

A million users is still less than an order of magnitude of change. The estimate moves from 12k/yr to 48k, quite low still.

I hope you can agree with my primary point, being that revenue share will not be the main source of income for successful creators, but rather the integration with other services that will require separate subscriptions/micropayments.

Similar to how it happened in the AppStore, but removing the option of ad rev.

I would say around 8-9 million from what I have heard from OAI staff.

It’s completely reasonable not to put all eggs in one basket but instead think ahead and look for other ways to monetize in case the revenue sharing mechanism does not cover all the costs.

The way how one can use a GPT to earn some income in a way that matches their business model, is exactly dependent on what the business is.
Personally I believe a freemium model is a good enough suggestion for many builders. But then we have to talk about the actual implementation of such a approach, which is doable but requires additional development.

What are your thoughts?

1 Like

I used PluginLab for plugins. They build a proxy that sits in the middle of the API backend and ChatGPT. It was easy to set a free quota and then charge beyond that. the same thing will still work with GPTs… it’s just that I think end user AKA customer expectations changed now with the GPT Store.

It depends mostly on the type of service you provide and your cost structure.

Freemium is nice if you think users will want to go beyond it, be it with more usage or more features. It is also only feasible if either your cost of acquisition and free runs is low or if you have the financial backing and conversion rates that justify the free service just to get a percentage of those as paying customers eventually.

Freemium models worked great in web2.0 where you could at least cover some costs with advertisement, but this is not really possible (mostly) when interating with your service through a GPT.

So, freemium models may need to become more restrictive than in the past, and at the same time increase their conversion rates. There will be some winners, but less and further between than in other 2-sided marketplaces.

If you’re gonna go with a freemium model, and you have enough resources to push through the early curve, I’d encourage to have it be full-featured but very limited on usage. You can also implement multiple payment models. Most new AI apps go with credits for low monthly subs, but I also like the pay-as-you-go of Dall-e 2 as a feasible option. I think we will see a rise of microtransactions in the GPT store.

  • User: Create a video from the images we just generated
  • GPT: "Oh you want another video made but run out of credits. Say ‘get more credits’ and we will load an extra 100 credits from your card on file "

:wink:

What other strategies have you seen or plan on exploring?

1 Like

Agreed. These micropayment credits could be handled by an aggregator service such that when a user tops up their prepaid account it can be used for any participating GPT.

Thanks for sharing your perspective on the problem!

I agree that the lack of “payments” inside the ChatGPT interface is the biggest question mark right now.

At the same time we know that OpenAI is trying to get more users for their ChatGPT service.

They incentivize developers to build GPTs but until now only with the promise of revenue sharing.

Thus if we drive new paying customers to their site, or if the paying customers return to use our GPTs then we potentially have a win-win situation.
This model would be similar to YouTube for example where content creators get paid to keep viewers on the platform.

From there we have to ask where the money to be shared will be coming from.
Obviously there is only a limited amount of Plus money to be distributed. Looking at YouTube we know how they fight to keep the platform profitable with many ads and without add blockers and extra transactions on top.

Following this line of thinking “revenue sharing” would be “ad revenue sharing” with a very, very low payout per transaction. But this is not going to drive builders to create GPTs for ChatGPT.

We have to expect that there will be a solution offered by OpenAI besides revenue sharing. And that would be a “paid action” feature that does exactly what is written on the box and drives earnings for both: the developer and OpenAI.

1 Like