Source item

I was a bit confused by the quote tweeted chart. Were there really that many AI companies making $100M in 3.5 years, that it was only a top quartile thing, given that ChatGPT is only three years old? I spoke to @jmelaskyriazi and he said his quartile graph is not an empirical cohort chart. That is, it’s not taking the revenue trajectories of 1000 AI companies and then showing that the top 250 hit $100M ARR in 3.5 years. It is instead an extrapolation of growth rates based on three buckets (he has a post on his methodology). Without meaning any offense, I think this extrapolation could be misleading. Because a true cohort chart that includes the countless AI startups that don’t ever make it to $1M ARR would show that getting to $100M ARR in 3.5 years is still amazing, and not merely a top 25% phenomenon. Probably at least top 5% if not top 1%, depending on what counts as an AI startup. That’s because the companies that quickly hit $100M ARR are typically ones whose names you know, like Midjourney and Cursor (see graph below). Moreover, Replit and OpenAI, impressive as they are, wouldn’t qualify because they were founded way more than three years ago. So: if you’re a founder, yes, getting to $100M ARR in 3.5 years is still phenomenal. And getting to $72M ARR in 5 years is also phenomenal, particularly if you can maintain strongly positive margins. Even with inflation, these are tough numbers to hit. Yes. It is possible to do something like this. But it is a top 1% or really <0.01% outcome. I disagree with this view for a few reasons. (1) First, fast topline growth businesses are rare. Often they have a thesis on how margins can improve. And sometimes that thesis plays out. AI costs are highly dynamic, for example. (2) Second, yes, from today’s vantage point we can see that ride-sharing worked out but scooters made less money. However, even Uber lost money for many years. And the micromobility companies did create enormous consumer value, and Lime is apparently going public. (3) Third, just about every major tech company was attacked on profitability grounds for years. Consider the headline below, or “Google’s toughest search is for a business model”, or the articles proclaiming Amazon dot bomb. (4) In short, it’s just not easy to determine which fast-growing companies will improve margins. Tesla did it relatively late in life. If you can do it well you could be a great growth stage VC.