Vibe accounting?

🤔 $10M ARR claimed, $3M actual – 11x.

Seemingly every week this year a new AI company announces they have smashed some record for getting to $1M MRR in X days/weeks/months with just 2 people.

Why? PR is back as a GTM motion, at least for AI companies.

This makes sense. AI is hot. AI might take all our jobs. AI might destroy us in service of creating more paperclips. Or something more fun and creative like blacking out the sun with a Dyson Sphere to satiate their unending thirst for energy, as suggested by Eliezer Yudkoski.

AI makes headlines.

So, if you’re an AI company, you’re growing fast. Make. Some. Noise. You will get headlines, and funding (with more headlines), and more growth!

The first wave of successful AI companies has been in areas where they are easy to deploy, lots of bottom-up adoption, and are generally low risk. Coding assistants, GPT type chat, AiSDRs. Hype + easy adoption has led to record-breaking growth. Some legit, some less so.

What's the catch?

Tools like Cursor or Windsurf, are clearly useful, but how long will it be until the next batch of YC launches 20 competitors?

In a related post, Oliver Molander relayed a rumor from YC W25 batch that a company was claiming “50M in annualized profit in 2 months.” As Oliver notes, there are standard metrics for this. While this company may well be doing great, it does feel like stretching to find a headline-worthy metric.

If we give them the most favorable interpretation, that they have made $8M in profit in those first 2 months, that is amazing, but it raises another specter: how could it possibly last? As Jeff Bezos famously said, “your margin is my opportunity.” If all it takes is 2 people and 2 months to create a company that solves this problem, those margins surely won’t last.

Competition is one challenge for these businesses, and the biggest risk I see for products like code editors that have shown they have persistent value, even if plagued with low switching costs.

Another bucket is tools like Lovable.dev, which is getting rave reviews for making it super easy to build an app or a website and achieved $17M ARR run rate in 3 months (again, incredibly impressive). But will users simply build the project they want and then leave? Ongoing usage seems as though it will require a shift in behavior. If websites are easy to build, maybe you will start building them all the time? But it’s not clear to me that will happen quickly.

A more likely scenario is that this will be more episodic rather than ongoing. This may not be the end of the world. Rand Fishkin, wrote an interesting contrarian take on this. He built his, non-ai business SparkToro with episodic use in mind and isn’t overly focused on the traditional churn metrics. Rand is also vocally not following the traditional VC playbook of swinging for the fences to create a hypergrowth company or die trying. This model won’t work so well when you’re on the treadmill trying to hit the milestones for the next fundraise.

We also have had an illustrative case this week of what happens when the hype comes crashing down. AI SDR company 11x made the news when it came to light that 11x had claimed $10m in ARR to investors but it turned out that much of that revenue had churned, and they are actually closer to $3m ARR. As an aside, I submitted a demo request to 11x months ago and never heard back, surely their AI could have gotten back to me? My opinion on the category, having reviewed others and received a lot of emails from them, is that there is clearly potential, but it is still very early, so it’s believable that folks have churned.

This all may seem negative, but to be clear, I am not an AI skeptic.

There is already tremendous value being delivered and incredible potential for much, much more. But there is some chaos inherent in change happening this rapidly (ref: “cursor deleted 4 months of work”).

I’m sure that we will ultimately see some of these AI companies demonstrate durable growth at record speed. Over the past 10 years or so we’ve seen B2B companies getting to the $100m mark faster and faster, even without AI, and notable, Cursor does top that list.

Google’s $32B acquisition of Wiz last week shows us that perhaps “$100M ARR isn’t cool, you know what is? $500M ARR.” Wiz hit $500M in just 4 years🤯. Deel is not far behind, hitting $500M this February, but they have been in the news for, ahem, other reasons recently.

Overall, it’s clear that you can build durable businesses faster than ever before. It may just be that AI companies are the next iteration of this trend, combined with an incredible amount of hype around AI means that we’ll see Cursor hit $500M in 3 years.

Will we see Sam Altman’s fabled $1B 1-person company? I’m not so sure about that, unless we are counting Taylor Swift as a 1-person company, in which case we already have a $2B 1-person company.

Warmly,

Paul Dudley