Hacker News Reader

I do wonder what the RoI is here

15M is like one month of salaries for 500 engineers, so like 40% of their whole eng. team

In Q3 2024 Google spent $7.2B on "Sales & Marketing Expenses". The data is not broken down more granularly. If 50% of that is pure "Marketing Expense" then you'll see that the scale of the market these players are competing for is worth $3.5B PER QUARTER in operating expenses to attract customers.

https://www.sec.gov/Archives/edgar/data/1652044/000165204424...

If you add a $15M one time capital expense to this then it doesn't even appear as a rounding error on this cash flow statement.

Acquiring a long term asset acting as a pillar of your marketing strategy then it begins to look like a very good deal.

Sure, you can have very sound theory to justify that

But how do we measure it? How can we tell that it was successful / good idea or not?

Even things like what is the difference between "chat.com" and "chat.ai"?

Would "chat.ai" for 1M be more efficient than "chat.com" for 15M?

Not to dispute the main point you are making, but that's assuming 360K salary. OpenAI engineers make significantly more than that.

Reply on news.ycombinator.com