The AI investing boom, or perhaps bubble, is something Silicon Valley has seen many times before: a gold rush of venture capital money thrown at the Big New Thing. But one aspect is completely unique to these times. Startups are rocketing from zero to as much as one hundred million dollars in annual recurring revenue, sometimes in a matter of months. Word on the street is that many venture capitalists will not even look at a startup that is not on this ARR superhighway, aiming for one hundred million in ARR before their Series A funding round.
Andreessen Horowitz general partner Jennifer Li, who helps oversee many of the firm’s most important AI companies, warns that some of this ARR mania is based on myths. She states that not all ARR is created equal, and not all growth is equal either. Li advises being especially skeptical of a founder announcing spectacular ARR numbers or growth in a tweet.
There is a legitimate, well-recognized term in accounting called annual recurring revenue, which refers to the annualized value of contracted, recurring subscription revenue. That is essentially a guaranteed level of revenue from customers on a contract. But what many founders are tweeting about is really revenue run rate, which involves taking whatever money was paid in a period of time and annualizing it. That is not the same. Li warns there is a lot of missing nuance about business quality, retention, and durability in that conversation.
A founder may have just had a killer month of sales, but not every month will necessarily repeat it. Or a startup may have many short-term customers doing pilot programs, so revenue is not guaranteed to stick around after the pilot ends. Normally, such boasts about growth via tweets should be treated for what they are. But because fast growth is a hallmark of AI startups, such claims are introducing a lot of anxiety to inexperienced founders who now ask how they can also instantly go from zero to one hundred million.
Li’s answer is that you do not. She says it is a great aspiration, but you do not have to build a business that way, only optimizing for top-line growth. A better way to think of it is how to grow sustainably, where once customers sign up, they stick around and expand their spend with your company. This can lead to growing five or ten times year-over-year, meaning growth from one million to between five and ten million in year one, to between twenty-five and fifty million in year two, and so on.
Li points out that this is still unheard of levels of growth. If it is coupled with happy customers and high retention rates, those startups will find investors willing to back them. Some portfolio companies in her group have hit those racing ARR numbers, like Cursor, ElevenLabs, and Fal.ai. But that growth is tied to durable businesses, with real reasons behind each of them.
Li also notes that kind of growth comes with its own set of operational problems like hiring. The challenge is how to hire not fast, but the right people who can jump into that type of speed and culture. The answer is not easily. That means the first one hundred employees wear many hats and missteps are bound to happen. Last year, for instance, Cursor angered its customer base with a poorly rolled out pricing change.
Li pointed out that other fast-growth startups deal with legal and compliance issues before they have systems in place, or face new AI-age issues like countering deepfakes. So while lightning-fast growth can be a good problem to have, it is also a little bit of be careful what you wish for.