1Mby1M Virtual Accelerator AI Investor Forum: Marina and Nick Davidov, DVC (Part 3)

Sramana Mitra: Let’s come back to the wrapper question. You started off by saying Perplexity is a wrapper on top of other LLMs.

Nick Davidov: Yes. Perplexity is actually very lean. If you compare how much money they raised, they haven’t spent much of it. They still have the majority in the bank. That’s their superpower.

OpenAI needs to spend $200 million to train a new model. It’s like the Hollywood blockbuster business—you may hit the box office, but the odds are low, and even if successful, your model might be relevant for only a few months before a new one replaces it. They need massive scale quickly to recoup the investment.

Perplexity doesn’t burn money on training compute. Their costs are minimal and directly tied to what users pay. Plus, they’re closer to the user, which allows them to control the flow of money better.

Sramana Mitra: So, here’s another related question. Lovable just raised a huge amount of money. It didn’t need to—it was growing very fast.

Nick Davidov: Was it? I don’t know what their profit margin is. Many coding products have negative margins.

Sramana Mitra: But Lovable is a paid product.

Nick Davidov: Yes, but in subscription-based products, some users burn more tokens than they pay for.

Sramana Mitra: That’s a pricing model issue. If the pricing model doesn’t support unit economics, that’s a business problem.

Nick Davidov: True. But take a look at Cursor or Windsurf—all of them operate at negative margins.

Sramana Mitra: That brings us to a very important point: Silicon Valley often builds businesses with poor unit economics—venture capital subsidized. We call it “venture welfare.” Why expect to build a sustainable company that way?

Nick Davidov: It’s about land grab. You’re looking for openings to penetrate the market. The belief is once you reach a threshold, you can use brand power, distribution, and funding to create a healthy business. Uber and Amazon are good examples.

Sramana Mitra: I disagree. Amazon lost money but didn’t operate with flawed unit economics. It invested in growth but always had a sound business model.

Nick Davidov: Where do you draw the line between buying growth and flawed unit economics?

Sramana Mitra: Let’s move forward. I only have eight minutes left.

Nick Davidov: I think this topic is important for founders. How do you decide whether to subsidize it?

Sramana Mitra: It is indeed a very important topic. In our 1Mby1M program, we emphasize bootstrapping first, raising money later, and maintaining profitable unit economics. That’s the only way first-time founders can build fundable businesses. When you came into this call to start with, you said

that you don’t invest in first-time founders. Our audience is largely first-time founders. And these first-time founders have nowhere

to go other than bootstrapping and building sustainably

and then getting to success. And then they can attract money.

They can attract money with a solid business. They cannot necessarily attract money upfront.

Nick Davidov: The funny thing is, most repeat VC-backed founders were once first-time founders who raised VC. I raised my first check when I was a first-time founder; so did Marina. So, it doesn’t mean they had to bootstrap. Many VC funds invest in first-time founders—it’s just that not all are ready.

Sramana Mitra: Exactly. So how do you get ready? You validate your business, build traction, and show fundability.

Nick Davidov: But you don’t need $1M ARR to raise VC.

Sramana Mitra: Many funds do ask for $1M ARR before writing checks.

Nick Davidov: Then raise from angels, not from big funds.

Sramana Mitra: We’ve been doing this for 15 years. We work with a large number of founders, and most are not fundable yet. They need to learn how to become fundable and make progress step by step.

This segment is part 3 in the series : 1Mby1M Virtual Accelerator AI Investor Forum: Marina and Nick Davidov, DVC
1 2 3