Why did Anthropic out-earn OpenAI with a fraction of the users?

Anthropic makes much more money than OpenAI with a fraction of the users: ChatGPT has over 1 billion monthly users, Claude about 245 million. OpenAI's early lead entrenched it. It kept building for the consumer virality it was born on, while Anthropic went after coding and business users, where the pain is high and so is willingness to pay. The lesson for any product: pick the problem with the most pain and the most pricing power, and build for the people who feel it.

How did Anthropic's revenue blow by OpenAI's even though GPT had a significant lead at the start? It is worth thinking about as an entrepreneur, and for what we can learn from it. In many examples the first-mover advantage holds and one becomes dominant, like Uber vs. Lyft. That was not the case here, so what made the difference?

Did Anthropic really make more money with fewer users?

Yes. ChatGPT has over 1 billion monthly users. Claude has about 245 million. Anthropic still makes much more money.

So Anthropic must have monetized better. That comes down to better conversion, higher willingness to pay, or both.

Why did OpenAI's lead hold it back?

OpenAI's initial lead entrenched it. It was born on consumer virality, and it leaned into that. Users asked for things like an image model or a video model, so it focused on an image model, a video model and the GPT marketplace.

Most products optimize for their initial customer. You also have to think strategically about who your ideal customer is and who you want to build the product on the back of. OpenAI was spread thin across those areas, and the problem it focused on was helping people get simple answers. That is what Google's business was built on, and it is a big opportunity. But there are a lot of cheap and free alternatives, so willingness to pay is low.

Microsoft's investment added a second kind of inertia. Azure had exclusive cloud rights, so if you ran on AWS or Google Cloud, you could not buy GPT there. That exclusivity ended in April 2026.

What did Anthropic see that OpenAI missed?

OpenAI missed the problem its product is naturally suited to, the one with a huge revenue price tag: coding. Anthropic saw it and went after it.

Anthropic focused on critical thinking and full explanations, where OpenAI was optimizing for quick, snappy responses that make you feel good. Of course, consumers don't like that. For business users, that depth of thinking was the product. I switched to Claude early for exactly this reason: I could see it would be better for my own business use.

Uber held its lead over Lyft for the same reason: it played to the strengths of the product, geographic density of both supply and demand, and did that better than anyone.

Why does willingness to pay start with the pain?

Willingness to pay tracks how strong the pain point is. Simple answers are a commoditized problem, so people will not pay much. Coding is painful, expensive in time and money, and common, so people and businesses will pay to solve it. Once you pick that problem, your go-to-market gets clearer, because you can see who pays.

What should a product team do with this?

I work a lot with products on defining their product thesis: who exactly your target customer is, what specific problem you are solving for them, and the two dimensions you can differentiate on better than anyone. Choosing the problem leads you to the people. Then list your top 5 to 10 audience segments. Grade each one on willingness to pay and on size. I bet the answer becomes clear. Then talk to your top 10% of users, the ones who love the product, and learn what they love. That is the Superhuman product-market fit (PMF) approach.

Both products have chosen a lane and a strength. The lesson is the power, and the revenue, of focusing on the cross section of high pain point and high pricing power. Here that was coding and business buyers.

FAQ

Did first-mover advantage fail for OpenAI? On revenue, yes. Its lead entrenched it in the consumer product it was born on, and that made the bigger opportunity harder to see.

What is the cross section of high pain point and high pricing power? A very painful problem that is not solved well and costs people a lot of time and money. For Anthropic that was coding and business buyers.

How do I find the segment with the highest willingness to pay? Grade your top 5 to 10 audience segments on willingness to pay and on size, and study your top 10% of users.

Michael Berliner advises founders and product leaders at consumer subscription apps on growth, pricing and product strategy through Otter Growth Advisory.