One key way AI businesses are different from the ‘old’ Silicon Valley
This is The Takeaway from today’s Morning Brief, which you can sign up to receive in your inbox every morning along with:
If the defining adage of the social media era was that you are the product, what new economic metaphors will the AI transition inspire?
A key difference between the new crop of AI startups and the mobile-era cohort of software companies is how they view the work and value of acquiring new users.
User growth was the coin of the realm for the past several decades, as sharply portrayed in a definitive scene from “Silicon Valley.” Why should app developers seek out revenue streams when they can simply run companies that are “pre-revenue?” After all, once you show revenue, investors will start judging your company’s financial performance. Better to chase the dream of a big payday in the future than worry about making modest money now.
The AI infrastructure investment train is running this playbook, in a way. Spending is the new growth. But something’s changed.
“AI-driven startups are building businesses that differ widely from the highly successful but now outdated ‘software as a service’ (SaaS) business model,” wrote DataTrek co-founder Nicholas Colas in a note to clients earlier this week. In other words, this isn’t your grandfather’s SaaS.
Growth is still king. But the economics of AI — that is, the enormous computing costs associated with training and running models — have forced the next generation of tech companies to reorient their sales strategy.
The barrier to becoming an “AI company” is incredibly low right now. Just ask anyone with the courage and entrepreneurial zeal to pursue home office vibe coding. But where the incremental revenue from every new customer was almost pure profit for software startups of the prior period, AI startups now face stiff computing costs, Colas pointed out.
Software companies were fixated on gobbling up users, adopting a mantra of growth at any cost, and lending their platforms in exchange for a fee. As Colas observed, the tech giants, too, structured their client relationships on a fixed-fee model.
In contrast, a new generation of AI startups has to focus on paid options from the get-go. Having the “right” users — as in, early adopters willing to fork over cash — is more important than pure growth. And, most importantly, their sales models will be proportional instead of flat because their computing costs are so high.
“This difference is one under-appreciated reason why AI valuations are so high,” Colas wrote. “These companies are deliberately pricing to shift profits from users to their own income statements.”
Leave a Comment
Your email address will not be published. Required fields are marked *