Atrium

» see atrium.co as it lived, on the wayback machine

~ the obit ~

Justin Kan announced Atrium's shutdown on 3 March 2020. Just over a hundred people lost their jobs, part of the $75.5 million the company had raised went back to the investors who had put it in, and the Series B lead, Andreessen Horowitz, took the write-off on a bet it had made eighteen months earlier. For a startup that had been one of the most-watched legal technology companies in the United States, the ending was unusually tidy and unusually honest.

Atrium started in 2017 with a diagnosis most startup founders would have signed. Corporate legal work for early companies is repetitive: incorporations, financing documents, employee option plans, the same forms with different names in them. Firms bill that work by the hour, which means the firm's revenue goes down when the work gets faster, so the work does not get faster. Atrium proposed to be a law firm and a software company at once, charging a flat subscription and keeping the difference as the software absorbed the repetition.

The structure was two entities: Atrium LTS, the venture-funded technology company, and Atrium LLP, an affiliated law firm staffed with actual attorneys. Clients were startups, the kind Kan knew by name. The archived site lists Aptible, MessageBird, SendBird, Triplebyte, ZeroCater and Volley among them, which is a credible book of business for a three-year-old firm.

~ the rap sheet ~

~ the pitch ~

What made Atrium interesting is that it employed the lawyers itself rather than selling document-automation tools to firms that employed them. That meant it could change how they worked by instruction rather than by persuasion, which is the step every previous attempt in this category had been stuck on. Kan's own explanation afterward was that this did not work either. His read was that full-stack legal models struggle because lawyers with established practices resist changing their workflow, and the partners who would have to mandate the change are generally the oldest people in the building. Owning the payroll turned out not to be the same as owning the habits.

~ the january pivot ~

In January 2020 Atrium laid off its in-house lawyers and said it would become a pure software company with better margins. A group of those lawyers immediately set up on their own and took clients with them, which tells you where the client relationships actually lived. Michel Narganes and Matthew Melville kept the Atrium law firm going as a standalone practice after the technology company was gone.

The pivot was supposed to buy time and instead burned the last of it. Customers who had bought a combined service were told mid-engagement that half of it was being removed, and some of them were unclear on who was representing them. Six weeks later the whole company closed.

~ what was left ~

atrium.co kept serving its marketing site for months after there was nothing behind it, last touched on 24 April 2020, still headlined "Exceptional professional services delivered differently". Today the domain answers with a 114-byte parking stub.

This is about as far from my own working life as a dead startup gets. I have never incorporated in Delaware and never signed a SAFE, and the whole category of problem Atrium existed to solve belongs to a funding culture that works differently where I live.

What travels is the shape of the failure, which has nothing to do with law. A team looked at a slow manual process, assumed software would compress it, and found that the slowness was load-bearing for everyone being paid by the hour.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »