

Discover more from One Year Wiser
Remotion and Alexander Embiricos
With a head start in 2019, Remotion launched right when the pandemic forced many companies to go remote. They enable remote teams to build culture through casual conversations.
The tech world was thrust into remote work in early 2020 and there’s been a steep learning curve to figure out how to work together and still feel like a team. Tons of companies and software tools have been built in the past two years to improve the remote work experience. Alexander Embricios and Charley Ho had a head start because they began working on remote work tools before the pandemic started.
Strong relationships (and trust) with your coworkers is essential to create a high performing team. But doing that in a remote only environment is much harder. Before I started working remotely for Reddit, I had the advantage of working in Reddit’s San Francisco office for two years. I had already built strong relationships with my team and stakeholders when I moved, which is a key component of a product manager’s success. But over the past two years more people have joined the company and I’ve never met many of my team members in person. We have had to be very purposeful in our interactions to build relationships and help foster a team that is trusting and supportive of each other.
I was excited to connect with Alexander to hear how they have been tackling this problem and how the product had shifted over the past two years. The launched Remotion 1.0 in March 2020 and the 2.0 version about a year ago in July 2021.
After he shut down his first startup, Alexander reconnected with Charley in San Francisco. They immediately clicked and knew they wanted to work together. They brainstormed ideas, identified where they wanted to focus and started to work, but before they got too far, Charley had to move to Chicago. All the people in the tech industry that they spoke to advised them not to continue working together: founding teams that are not based in the same city never work.
With this flood of advice about how hard it is to collaborate remotely, they decided to pivot their startup to focus on enabling teams to do just that, collaborate remotely and still build culture. This was in late 2019. There were some companies that were remote first, but they were few and far between. Alexander had conviction that remote work would grow, little did he know that a global pandemic was about to accelerate that trend.
Remotion helps teams build culture. The first version of the product focused on enabling quick video chats between team members, but they quickly pivoted in 2020 when society began to experience “zoom fatigue,” and more video calling was the last thing employees wanted. Now, Remotion is a virtual office that facilitates casual conversations with coworking and shared music.
In our conversation, Alexander talks about:
How his background, not the pandemic, made him interested in enabling remote work
The mistakes they made with their initial launch
How they recognized the need to pivot the product as “zoom fatigue” set in
The importance of building a product that is opinionated about how people should use it
How he handled the challenges of purposefully deciding not to grow and instead focus on foundational pieces of the product
My questions are in bold; this interview has been lightly edited for clarity.
Let’s start at the beginning. What’s the founding story of Remotion?
Alexander Embiricos: A few years ago, I started a gaming company in San Francisco with two other co-founders. We had limited savings living in an expensive city. Six months in we raised $2.5m dollars from First Round Capital for an unusual purpose: Founder salaries and no hires. Looking back, I’d call it both an abundance of caution and simply being nervous about product market fit. Either way, after over a year of grinding, we churned through a few ideas without building conviction around any. And more importantly, we’d realized that we didn’t want to be gaming founders. We ended up shutting things down. Fortunately, we’d only burnt $300k.
First Round Capital was supportive of me continuing to try ideas using the original capital. I was spending time thinking about new ideas when I reconnected with my co-founder, Charley Ho, who I had known in college. We were both living in San Francisco. We quickly decided we wanted to work together on a business. We were very aligned on values and that we wanted to build a B2B business where customers directly paid for value—not an ads business where customers are actually the product.
We didn’t have to pitch anyone because we still had some of the original capital. Which was good because honestly, I was nervous that no one would fund me after my first failure.
But, shortly after we’d started working on a web-based RPA tool, Charley had to move from SF to Chicago. We were concerned about starting a company from different cities. We began asking other founders and investors for advice on co-founding remotely. All the advice was don’t co-found remotely. Maybe you can hire some remote engineers, but definitely don’t live in a different city from your co-founder. But we just really wanted to work together. So we decided to give it a try.
Remotion started as a tool we built for ourselves. As we were exploring how to make our remote-first company work, we realized that most of the best practices assumed asynchronous work. Casual conversation was treated as both impossible and bad. Since we ourselves like hanging out remotely, we built a tool to enable these casual conversations and we ended up being more excited about working on that than the other B2B SaaS automation tool we were building.
We switched to focus 100% on Remotion a few months before COVID hit.
Wow, that’s very fortuitous timing that gave you a headstart on all the startups that pivoted to enabling remote work in March 2020. Were you primarily focused on enabling remote work because you and your co-founder happened to be in different cities or was there something else you were seeing in how Silicon Valley was shifting?
AE: Personally, I was very ready for remote work because of my background. I'm half Greek and half Malaysian. I moved to America for college and as a new grad, I wanted to eventually leave the Bay Area and move back to Europe, but it didn’t feel feasible to me because I had this high paying tech job at Dropbox and no way to do it remotely.
I always felt that it'd be better if products were worked on by teams based all around the world and I thought it was strange that it wasn’t already happening, given the nature of software development.
So it was this background that made me excited to jump on the idea of Remotion as soon as we realized there was a problem of creating casual conversations between remote teams.
When was that the initial launch of the product and how were you gauging success of the launch?
AE: To be honest, I didn’t realize we were launching. We had started working on Remotion a few months before the pandemic hit. We were in it for the long haul. We believed that remote work was going to become more popular, but that it would take a long time… and then the pandemic hit and everyone was working remotely overnight.
We didn't feel like our product was ready and we didn't know what a launch was, but we thought we couldn’t pass on timing. We put a self-serve flow together over a few days so that people could sign up for the product and then we posted it to Product Hunt. We positioned it as something that is new and not quite ready, but that we wanted to be helpful to teams that were struggling to work remotely all of a sudden.
But, I would do that so differently. If I could do that again, I'd be much more intentional about it.
Also, we had made some really weird decisions for the first version of the product, like building it as a native Mac OS app, which resulted in us having slower releases and not having cross platform support.
After the launch, we weren’t really focused on acquiring users or any specific metric because we had no idea what to expect. I’m shocked we got as far as we did with our “build it and they will come” mentality. But feedback around bugs began to pour in like crazy and we were just firefighting.
The product was rough and not ready for prime time overall, but we saw a small group of teams that had just crazy usage stats and that were growing over time. Also, people kept writing in with the word “love” — which I hadn’t seen that often at Dropbox. People were jumping in and out of tons of conversations in Remotion. That handful of “Remotion-native teams” is what gave us the conviction to dig in and raise enough money to weather COVID.
So you launch a simple Mac App in April 2020 to help remote teams facilitate casual conversations, find some traction, and start improving the product. About a year later, in July 2021 you launch 2.0 on Product Hunt. How did your product and your strategy change from 1.0 to 2.0?
AE: We'd always had this vision of building a product for teams “who like hanging out,” but started with something we thought was easier to explain: “fast video chat” for easier collaboration. The solution was lightweight. A floating “dock” on your screen; you can click on a team member and start talking with them, sort of modeled off the in-office behavior of walking up to your colleague’s desk and tapping them on the shoulder.
We were using a similar framework to Superhuman where we asked users how disappointed they would be if they stopped using our product. We had a high percentage of team leaders who would be very disappointed if our product didn't exist, but we found that many people on their teams didn’t care, and were not adopting Remotion.
We realized we had two different audiences with different needs. There’s the team leader who has a bunch of incentives to build culture, team cohesion, and productivity. Then there’s the individual contributor, the IC, who is overloaded with work and tools. Our product actually needed to serve them both and we weren't managing to consistently get the second group on board with our product.
Around this same time, “zoom fatigue” was starting and we watched our growth flatten and then start to go down. The market was changing.
As we spoke to customers we realized the “quick video chat” was becoming an anti-value prop. ICs didn’t want more video calls. But, we learned that teams were still using it for casual hangouts. It was like they had squinted at our product, ignored all of the text on our landing page, and somehow saw our initial vision of casual conversations buried within the product.
I’m sure watching the market change and become anti-video calls was a stressful time. When you realized that your customers were using it for the casual side of team building did you change the product significantly?
AE: Yes, and we made some unusual decisions to get there. First, we explicitly decided not to grow and closed sign ups. (When we had launched the product had open signups—anyone could sign up and begin using the product.) We got rid of that and introduced a waitlist. Instead, we asked people who signed up why they wanted to use the product, what they needed, what they wanted.
These interviews helped us evolve our product strategy and it shifted significantly from 1.0 to 2.0. Instead of quick video calls, we were now focused on helping remote teams build a better culture. We weren't exactly sure how we would help teams with culture building, but we knew vaguely that that was the problem we wanted to focus on.
Instead of building everything out, we decided to do a 2.0 launch on Product Hunt with our new positioning and value props, plus a couple of new features, to increase our waitlist of people willing to be interviewed by us.
One of the big changes was that we redesigned the product to be anchored to specific intents at specific times. This was better for managers seeking adoption and better for introverts. For example, we previously had a feature where you could set yourself as “Open to talk” to allow teammates to start calls with you without needing confirmation. With 2.0, we replaced that with “rooms.” Different rooms serve different purposes. As we iterate on how to encourage hanging out in rooms we’re adding music and small games that differ depending on the room and its purpose.
Sounds like switching to a private beta was the right call. Do you wish you had launched the first version of your product with a private beta?
AE: Knowing what I know now, yes I would have. We had so many bugs at the beginning due to all the various machine and OS configurations. We got sidetracked by all of that instead of focusing on whether the product was providing value to remote teams. A private beta would have just let us control that more and ensure that we learn more with each new person that signed up.
Do you think you’ll eventually switch back and open up signups?
AE: Yes. Right now we are working on our website to make it more specific, and then we plan to switch to a self-serve onboarding depending on which landing page you arrive on. For example, if you come to the landing page that is all about Spotify and listening with your team, we don’t need to coach you on how to use the product when you just want to listen to Spotify together.
The reason we're not changing to that overnight is because the activation rate for someone who we onboard with a call is double the self serve. Obviously there's some selection bias because someone willing to schedule an onboarding call with us is going to have higher intention. But there's still an element that I’m concerned about which is that we are building too broad of a product that is only vaguely useful. We recognised that and we're building a product that is more opinionated about habits and how people should use it.
Going back to how you shifted the product from quick video calls to collaborative rooms for casual conversations. How did you manage to stay open about the direction of the product rather than sticking with the original idea even if adoption was slowing?
AE: It’s a couple things. First, it’s talking to users. It’s so critical. It's nothing novel, but we've gone through extreme lengths to enable that. As I mentioned, we had a waitlist where we talked to every single person that signed up for our product.
Having gone through that experience where we were building something which turned out not to be what people wanted, we pay really close attention to the feedback from our users. The switch from remote only teams to hybrid teams is a good example of this. As more and more people begin to go back to an office for a few days a week we know that our product needs to evolve. Coming soon: product updates to support hybrid teams.
Related to feedback, we’ve been much more specific about which customer should use Remotion. When we first launched, we said anyone who works remotely can benefit from our product. We wanted the market to be nearly infinite. But, now we're much more specific. We're targeting engineering leaders at companies sized 11 to 50, who have a mostly remote team, their career page includes the words “hang out,” and ideally they are hiring because hiring means onboarding and onboarding is a good time to use Remotion.
Being super specific about who should use our product makes it a lot more tractable to hear and digest the feedback from customers. It’s less scattered. We aren’t distracted by feedback from someone trying to use our product for their classroom. Focusing on a small group of people first is something that you read about in business books, but at least in my case, I had to make the mistake first before I became truly disciplined around it.
Lastly, we’ve also been much more intentional about breaking down different stages of the activation funnel. How does that look for a team leader who signs up for our product? How does it look for the team members they invite to join? By tracking these metrics we can measure whether changes in our product improve activation rates or if something breaks in the funnel.
Focusing on a small group first sounds like it was a hard earned lesson for you. Thanks for sharing it with me. So, it’s been almost one year since your 2.0 launch. Where are you hoping to be a year from now?
AE: I view last year as our iteration year. We used a Product Hunt launch as a way to get leads that we could learn from and iterate with. We didn't actually have a product that was fully there when we launched. Now we have a product that is working and has traction. Our goals for the upcoming year are to grow. We plan to open signups with maybe optional demo requests, remove the “beta” tag, and add support for more platforms. We are also going to focus on the increase in hybrid teams and maybe integrate our product into physical office spaces.
Growth is always a great goal! Let’s go to the quick fire round. What are you most proud of from the past year?
AE: That's easy. The team. I think we have an incredible team with strong talent and an incredible learning culture. I think we got really lucky with a few key hires at the beginning who have really helped us create a strong culture of learning. Also, I’m proud that my co-founder and I have had introspection around what we’re good at and where we can learn from others and as a result have evolved a little bit past the myth of the founder that does everything. We have people who are better than us at what they do, and that's awesome.
Was that something you had to learn how to do or did you always have a low ego?
AE: It’s not really an ego thing. It was more about learning that it’s okay to hire someone even when you haven’t figured it out yourself. It's almost like we thought so little of ourselves that we thought: “We don't deserve to have someone focused on growth because we haven't even grown enough by ourselves.” It's this weird anti-ego thing that is super counterproductive. I think it's important that it gets talked about because a lot of what you read about startups suggests that you need to be crushing it before you hire someone. We don't talk as much about the importance of identifying a problem in your business that you are struggling to fix, and then bringing in someone who is an expert and can help you with that problem.
That’s great advice. What's been the most challenging part of the past year?
AE: The most challenging part has been not focusing on growth. Going back to the drawing board to get feedback at the expense of signups, and taking the time to invest in some foundational projects while putting off work like Windows support.
Putting up the waitlist list was the right decision, because it helped us gather a lot of feedback from prospective customers. But it hurt our top line growth numbers and that hurt team morale. There's a lot of responsibility as a leader of the team, so I had to spend a lot of time thinking about team morale and strategy.
I’m happy to be refocusing on growth, but it’s still painful starting from scratch.
How did you tackle that as a leader?
AE: So instead of pointing to something like week over week growth in weekly active users, I had to constantly beat the drum of progress we were making by learning from the new customers. All of the insights we were gathering were going to help us make a better product.
But it was really hard. For example we were a video platform and “video minutes” was a key metric, but we were moving the product away from video calls so that metric was decreasing. Engineers were looking at those stats, it was their job to monitor usage and performance, and they were being confronted by decreasing numbers. It was a big growth experience for me to learn how to refocus the team on our new product strategy.
What book do you recommend the most and why?
AE: Extreme Ownership by Jocko Willink. It’s a bit of an annoying read, but was critical for me when I was a new Product Manager fresh out of school. It helped me understand my new (very strange) role as someone who led teams and was responsible for outcomes, even when I wasn’t doing the work myself.
I think product management is particularly hard for new grads because a lot of what you learn in school emphasizes individual contributorship over leadership and teamwork. There was this one project I was pitching at my company that I didn’t get resourcing for. So I decided to build it myself and I was super proud when I told my manager we’d shipped—at scale. But instead of congratulations, he gave me a lecture and that book. He said that going off and building the idea by myself wasn’t leadership. It was hard for me to understand at first. I had a very binary point of view on work, and this book helped me understand that leadership was amplifying a team outcome rather than just managing your own work output.
What do you know now that you wish you'd known a year ago?
AE: I actually have three things.
First, my general learning from the past year was how it’s absolutely critical to break your businesses top level goals into tractable components. Looking at a graph trying to read tea leaves is stressful and unproductive. Once you start breaking it down into different components, you can start to move them. It makes a lot more sense and you know where to focus. And that focus is key.
The second learning I had is that marketing is learnable and it needs early investment.
My last learning is for founders in the team productivity space: If a product isn’t shared, it’s dead. Sharing’s easiest when the product’s built around a specific intent at an acute moment in time. So now, when we think about new features, we always try to build around that. We're not trying to build something generally useful at some point in the future. It has to be useful for a specific moment in time and intent.
My last question for you, what advice do you have for a product manager that is thinking of starting a company in 2022?
AE: Ok, this advice comes out of my failure as a gaming founder:
Make sure you care about the people you’re serving, and enjoy spending time talking to them about their problems.
There’ll be highs and lows as you build your business. You might even realize that the solution you chose or the problem you’re solving aren’t worth it. But if you get energy from your customers, you’ll have the most underrated startup superpower: The ability to keep going.
--------
I hope you enjoyed my interview with Alexander and learned a few things. If you did, give it some love on Twitter with a RT or Like. This helps more people discover my newsletter 😊 -Tyler