If you’ve got 99 problems, not shipping is 98 of them

In late 2022, Perplexity launched its first product, Bird SQL, which let you search through Twitter data using natural language. Yesterday, they launched an AI-first browser, Comet. In the meanwhile, they launched the product that Perplexity is today – but they did 100 other things: a perplexity bot on X, a financial research agent, a search API, perplexity labs, enterprise product for internal search, a mobile assistant, a contest during Super Bowl, acquired multiple companies, and a lot of other things I’m not aware of. They also raised close to $1b which made a lot of this happen.

Some of these products, like Deep Research, sucked versus the competition. When they started Bird SQL, I’m not sure they clearly knew what the end goal was, what would actually succeed as a product. Perplexity’s strategy was not to find the right product, it was to ship many things as fast as they can, see what sticks, and keep trying. Momentum is what they were betting on. Momentum – by gaining users and going viral, is what let them raise money, acquire companies, build multiple products, and survive to see a new day and a new product get out.

Momentum matters. When you’ve got momentum, you’ve got users giving feedback, complaining – investors ready to give you money, a team with the morale boost to keep pushing forward. When there’s no momentum, you don’t have feedback from reality, so you don’t have a promising way forward, which hurts morale. The point I’m trying to make is this: it does not matter if the product that you build today is the product that your company will ultimately be selling. Instead, all that matters is – are you shipping fast, are you learning? Because if you aren’t shipping fast and learning – then you’re just delaying the inevitable death.

Tesla’s goal was to build cheap electric cars but they started with luxury cars. Their portfolio is much bigger today, and the vast majority of future revenue will come from a completely different market – autonomy. Their mission went from sustainable transportation to sustainable energy, and now it’s sustainable abundance, and Elon estimates that the market captured by Optimus will dwarf their current market.

SpaceX’s goal was to get to Mars, but I don’t think they could’ve known ahead of time that it would be a satellite constellation providing internet (Starlink) that will help fund their R&D.

The point is – they couldn’t have planned these pivots. They discovered them by shipping products, getting market feedback, and following the momentum wherever it led.

The core mission may or may not be the same – but the path you expect to progress on will almost certainly turn out to be different from the path you end up on.

So – the product you’re building today may not be the right form factor, it may not blow up, it may not even get off the ground – but unless you ship fast, get feedback, you can’t iterate. If you can’t continuously iterate, you won’t even know when to stop and move on, so you can only guess. And if just guess when to stop/iterate without shipping or feedback, then you’re back at square one.