Jason Rosenthal, behind Lytro's pivot
How this founder saved his company from bankruptcy and later sold it to Google.
Written by Immad Akhund
Our first-ever print edition of Meridian was themed “pivot,” dedicated to all of the times you’ve heard old startup ideas and said “they used to do what?” For the occasion, Immad Akhund, CEO of Mercury, had a fireside chat with Jason Rosenthal, operating partner at a16z crypto. Rosenthal talked about the difficult decisions that went into pivoting the successful light field camera company, Lytro, in 2013 — including finding a new market, navigating layoffs, communicating with investors, and eventually selling the business.
From all the times you’ve been at a company or led a pivot, what’s one that particularly comes to mind? Was it painful?
I think every pivot is painful. It’s difficult to realize that the thing you’re pouring your heart and soul into building is not fundamentally working. The hardest one was the last company that I was a part of.
The company was called Lytro, and I sold it to Google. We were a computational company with really revolutionary hardware and software. We wanted to change how all photography and video works. Some context: Historically, since the invention of photography, we’ve only ever been able to capture two properties of light — brightness and color. That’s the same whether it’s film, digital, still video, whatever. That’s what forms all of the pictures we see today.
With the help of over a decade of research at Stanford, Lytro invented a novel photographic sensor known as a lightfield sensor. In addition to color and brightness, you could capture all of the rays of light, all of the geometry around every single ray of light flowing in every direction. What that allowed you to do was apply everything that makes 3D computer graphics work — like visual effects in movies or video games — to photography.
The company started by trying to build a consumer-focused camera that let you take lightfield pictures. You could refocus a picture you took. It was really cool and computationally complex — it was also a horrible business and a really difficult product to build.
We built the first generation product in about three years, and shipped and sold over 10 million units. That’s not bad for a hardware startup.
Unfortunately, two other things were happening. The first was that we shipped our first product right as smartphones were getting equipped with really good cameras. We thought we were going to replace the traditional SLRs and point-and-shoot cameras. But the first Lytro camera shipped in 2012. And it was in 2011 that the iPhone camera first got good. Right as we were executing our path, this big disruptive force came in. It turns out that if the camera in your pocket is good enough, that’s the one that you’re going to use. It sounds obvious in 2023. But in 2012, we were still figuring it out.
In 2013, I took over from the founder and CEO, who went back to academia. Right after the first product, we shipped our second one, which was better, faster, and higher resolution. I had no choice but to see that through. We shipped it in 2014.
But it was quickly becoming clear to me that if we stayed on that path, we’d 100% go bankrupt. What we decided to do — we still had belief in the capabilities of the core technology — was to pivot. We pivoted the company to focus on the high end of the market. And we went after two markets for the same product. The first was the highest-end film and TV production. Think Marvel films, Star Wars, Game of Thrones. At the same time, we went for an emerging market of high-end captures for cognitive reality and VR.
It was a tough pivot. We started by laying off 73% of the company. We were a couple of hundred people, but we were burning cash like crazy. Everybody there was amazing and knew a lot about building cameras. Unfortunately, building a professional product is different than a consumer product. Everyone who worked on the first two generations of the consumer product was laid off.
I realized over Christmas at the end of 2014 that that was what we would have to do. It was the worst vacation I’ve ever taken.
Eventually, we got the interest and attention of all the top directors and cinema photographers. We were used in a number of big films and created a lot of the content for Oculus, which was new then.
In 2018, we had an offer to be acquired by a big company rumored to be releasing a new VR headset the following week. And then they ended up selling to Google. So our camera tech, patents, research is now used across Google camera systems, from Pixel to elsewhere.
Never miss another untold story.
Subscribe for stories, giveaways, event invites, & more.
I often talk to companies with [the] kinds of pivots, who say, “Oh, we have the tech, let’s go apply it to something new” [and I tell them] that’s a bad idea, and tell them to go back to the drawing board and find the best possible idea they can think of, rather than an adjacent one.
What do you think of that?
It's super situationally specific, right? When you’re building something, a lot of days are spent in the fog of war. If you knew what to do or where the market was or how to build a winning product, you would have done that. One of the things that makes doing a startup so exciting is that you’re constantly learning, you’re searching for product-market fit. I think you’re right that going back to first principles and starting with a net new thing is often better. You have a better opportunity to build something more successful. But that’s not always a move you can make. It depends on your team, your capital, your competitive dynamics in markets.
The other thing is that startups are often about change. You talked a bit about layoffs. How did you communicate change and get employees excited rather than scared?
You need to have deep conviction, a good story, and hopefully some amount of data about why the new thing is better than the old thing. In Lytro’s case, we decided to go to the part of the market we went to because we researched a ton. We talked to everyone we could — scientific microscope companies, medical imaging, film and television, all the different markets for imaging.
And when we got to Hollywood, we found that the most creative people in industry would not only complete our sentences about what they would do if they had a camera with the capabilities we could bring, but they would also come up with 5, 10, 15 new ideas that they thought would be really game changing if applied to their craft. So that was the first kernel that told us — hey, maybe there's something here that we should be doing after.
Do you find that every time you have been involved in a pivot, the writing's been on the wall for much longer? Is it often that you’re like, “I regret not doing that a year earlier”?
Absolutely. Andy Grove, the Intel CEO who executed arguably the biggest and best pivot of all time, going from memory chip to microprocessors, said that you can’t fake momentum. I didn’t understand this at the time, but when you’re leading a company, deep, deep in your heart, you know if the thing you’re doing is working. You can tell yourself stories like — it’ll be great if we just have these features, or more capital, or spend more on marketing. But most of the time that isn’t true.
But there are plenty of stories where it took two, three years to get there. Figma, Airtable, you know. I think that’s something that can be confusing. There’s the one story, where you shouldn’t give up on your work and you should persevere. But the other is that you should have given up.
What do you think is a good try before you consider a pivot?
Look, I 1000% agree with you on not giving up. But where I want you to be clear with yourself is what working looks like. You don’t need to have 100K or 500M customers on day one. I’m in crypto these days, and there aren’t even a million active crypto users yet. You want to be able to know that the people using your product are getting real value out of it. They’re enjoying it, they’re coming back from more. You can then ask yourself how many more customers like that there are out there. And then you need to find them.
Or is it that the total size of the market is just smaller? Are people not loving your product?
How do you think about communicating with investors? Like, with Lytro, you guys raised a ton of money, and I think you did a successful raise after the pivot. How do you work with them through this journey?
I think the best thing is to be direct and honest. But also, remember — once an investor writes a check for you, they’re in for the ride. You can always return capital, but most firms are looking for a 1000X upside. I think the best thing you can do is walk your investors through the logic that you went through to pivot and answer all of their questions. And the best investors will have questions that you didn’t get a chance to ask.
More Like This