2021-05-28 ☼ memo
Justin Kan, the founder of Twitch which sold for $970M to Amazon in 2014, recently shared his story of now shut-down legal tech company Atrium.
Atrium was one of Silicon Valley’s biggest potentials. It raised a total of $75M from top investors like Andreessen Horowitz and General Catalyst. The legal services industry is broken, and Atrium was supposed to fix it.
But it couldn’t. Instead, Justin Kan announced Atrium’s closure before the COVID-19 pandemic in March 2020.
People love talking about success stories. Especially the Silicon Valley ones. Ones that go onto become the largest companies the world has ever seen. But behind the scenes, there are many more stories about failures, and we don’t hear them as much.
I have much respect for him, for sharing his failures. It is tough to swallow all the bitterness that comes with admitting and sharing your own mistakes and failures. Thank you, Justin, for the opportunity to learn from your experience.
My startup Hyperinbox (formerly Pixelic), was an Atrium customer. We loved the service. When Justin announced it was going to shut the company, we were truly disappointed. The fees for legal service were expensive for a scrappy startup that’s trying to stand on its own feet. Atrium was perfect for companies like us.
Successful stories teach important lessons, and they’re all great. But learning about failures is equally important. You learn just as much, and it motivates you as much.
Here are the links to his video, blog post, and Twitter thread.
And my summary of why Atrium failed:
Atrium raised a $10M Series A with an idea — a 10-slide pitch deck.
They had a great team, great investors, and even early customers that believed in Atrium.
Atrium was Justin’s product of “thinking big.” Insanely large numbers. A multi-billion dollar, world-dominating company. But was Atrium and what it’s set off to solve something Justin truly cared about and loved?
Quoting Justin directly, It is very hard to write the mission after the fact. You should start with a clear reason to exist and filter early hires for believers.
Atrium also hired too fast. It was unable to establish a cohesive culture.
This one resonated with me the most. Atrium’s idea was great. And their team was amazing too. They had the money to scale early on – which they did. But, at the core, Atrium didn’t have a product that people enjoyed using.
Atrium successfully grew the customer base quickly, but it had failed to retain them.
“True technology startups is about finding a differentiated solution using technology to build something that is fundamentally new.” – Justin Kan
Marketing and distribution as a primary differentiator are not good enough – if the product is not better. Customers come in, and if the product is not good enough, they churn out. This is what happened to Atrium. They had lots of new customers. They came to Atrium, but they didn’t stay. Justin spent six months trying to solve the churn problem. He couldn’t.
It was difficult to define who Atrium should serve – the lawyers or the clients. Atrium tried to serve everything to everyone. Didn’t work.
Twitch, on the other hand, defined its primary customer early on, the streamers. Everything Twitch did was prioritized for the streamers, not the audience. The audience comes to Twitch for the content (generated by the streamers), not the platform.
Justin had a “win or die” leadership style. It’s either you contribute to the team, or you don’t. It’s either you add value, or you don’t matter to the team. He now concludes, “a more empathetic approach would have at least been a morale boost for the team.”
Similar to #2. He realized his intrinsic motivation didn’t align with Atrium. He loved talking to interesting people, learning new things, and creating stories — all part of a CEO’s job description, but he didn’t love being the CEO or building products.
Justin also had no passion or genuine interest in legal tech. It’s important to build something that you have genuine interest in. Otherwise, you have little incentive to continue even during the hardest times.