If you happen to be the CEO of a company that makes foundation LLMs, this article may not apply to you. Your company is selling a dream that your models will be the way all work is done in the future.
For the rest of us, we still need to sell a business model, and to do that, we still need to rely on fundamentals that show how we can make money and how we can grow.
How Valuations Work
When trying to establish the value of the company, there’s no firm formula. Typically, established companies are valued more or less based on their current profits1, and early stage companies are valued more or less based on the perceived potential of that company to make profits in the future.2
With that logic, big AI companies like OpenAI and Anthropic would seem extremely overvalued (The Economist goes so far as to call it “verging on the unhinged”). Despite being huge and not that new, these companies argue that the promise of continued AI breakthroughs (that only they can achieve) justifies astronomical valuations. Whether or not this is true, enough investors agree with it that these companies can command enormous valuations.
Will they achieve super-intelligence before investor apatites run out, or find some other way to radically improve their profits? I don’t know, but regardless, these companies are outside the realm of how valuations normally work for a company at their stage – at least outside of a bubble.
Fundamentals still matter
For the rest of us, it’s important not to be taken in by this kind of thinking. Despite claims like “ARR is Dead”3, companies that are not AI Giants still need to focus on the fundamentals.
What are the fundamentals? These include basic metrics like ARR and churn rate for startups and profit and EBITDA at established companies.
Why? Because you probably can’t convince investors that your product is destined to produce trillions of dollars in the future with just a little more development, so instead you need to show them that you are making money now, and you have a clear plan to grow.
Even so, you are competing for investor dollars with AI companies with wild promises, so it’s all the more important you have good fundamentals. And of course, it’s not just about proving yourself to investors, but it’s also about proving your business to yourself. Do you know what your strengths and weaknesses are? Do you know where you have the most opportunities and risks? Only your traditional business fundamentals can tell you that.
Footnotes
-
The “Price-to-Earnings Ratio” is the number you’ll look for when evaluating a company this way. Savvy investors will discount future earnings because they are less certain and are worth less the further in the future they are expected, so a company expected to earn a lot in the distant future should be worth less than a company that is expected to make lots of money in the near future. Some mature companies trade higher than others with the same PE because investors believe they still have a lot of room for growth. You may not believe in the continued growth potential of companies like Netflix, Microsoft, Google and Meta, but the market seems to think they still have plenty of room to grow. ↩
-
Some amount of revenue of course validates a startup’s predictions by showing that someone is willing to pay something for their offering. Not that long ago, consumer apps didn’t even need revenue to earn investment – they just needed to show that they could engage and retain users. The idea was that once you had engaged users you could find a way to make money off them, which was often true. ↩
-
Of course it’s a bit of a click-bait-y title. The article itself is nicely nuanced about how changing market conditions mean we need to rethink exactly which metric we use to track the success of our business, and if we are going to use ARR, we should be careful how we define it. ↩
