My take on the AI startups, and in particular the generative app startups, is that those ARR numbers are a mirage. I personally signed up for Cursor and Replit to test them out. And perplexity and open AI. And I tried out stack blitz and srcbook and lovable and v0.
I'm not going to keep paying for all of them. I did it to learn, and now that I know what's capable, I'll keep looking at the releases, but I'll stick with one and keep switching to the one that's better than the others.
There's no moat, and there's no "ARR". Nothing is recurring about it.
I'm willing to bet that a good half of that revenue is one-off purchases (by behavior), annualized. Churn is going to be through the roof.
Let's see how they're doing in years 3-5. Any growth there is just going to be combating wild churn. M&A is a likely strategy that will condense folks down into a few key players.
I am also curious about the stickiness of these users, and some of these startups are going to face some huge churn concerns.
But there are reasons to believe those who have captured the dollars will continue to do so:
1. Don't underestimate laziness (i.e. people just leaving their subs on auto-pilot).
2. The weight of first mover advantage, especially if those first movers are landing business/enterprise clients. Unless a new boss comes in or an attempt to move AI functionality in-house becomes a priority, they are a relatively sticky bunch.
3. How quickly these AI businesses can compound around the behavior of their users to further customize and improve individual experiences. Then it's a matter of "firing" the AI that works for you now in the hopes that the next one is better — similar to how letting go of a low performer on a team can take a while.
My take on the AI startups, and in particular the generative app startups, is that those ARR numbers are a mirage. I personally signed up for Cursor and Replit to test them out. And perplexity and open AI. And I tried out stack blitz and srcbook and lovable and v0.
I'm not going to keep paying for all of them. I did it to learn, and now that I know what's capable, I'll keep looking at the releases, but I'll stick with one and keep switching to the one that's better than the others.
There's no moat, and there's no "ARR". Nothing is recurring about it.
I'm willing to bet that a good half of that revenue is one-off purchases (by behavior), annualized. Churn is going to be through the roof.
Let's see how they're doing in years 3-5. Any growth there is just going to be combating wild churn. M&A is a likely strategy that will condense folks down into a few key players.
I am also curious about the stickiness of these users, and some of these startups are going to face some huge churn concerns.
But there are reasons to believe those who have captured the dollars will continue to do so:
1. Don't underestimate laziness (i.e. people just leaving their subs on auto-pilot).
2. The weight of first mover advantage, especially if those first movers are landing business/enterprise clients. Unless a new boss comes in or an attempt to move AI functionality in-house becomes a priority, they are a relatively sticky bunch.
3. How quickly these AI businesses can compound around the behavior of their users to further customize and improve individual experiences. Then it's a matter of "firing" the AI that works for you now in the hopes that the next one is better — similar to how letting go of a low performer on a team can take a while.