I've been running a weekly happy hour with the Orbital members for the past few weeks, and it's been a nice ritual to see a bunch of familiar faces while simultaneously creating space to check in with each other.

If there's been a downside it's that we've had to use Zoom, because it by far provides the most consistent, scalable performance in terms of both audio and video quality.

The challenge is that that Zoom has always clearly been a utility but it's become a social platform due to the pandemic–it's the primary vehicle by which we engage with other people.  And so, as a result it's inherited the mental models that come with social platforms:

  • it's primarily associated with work, so there's a dissonance when it's used for social interactions
  • the user experience of the product is designed for business meetings, and so it forces you to run your social experiences like a meeting. You'll need someone to lead and facilitate.
  • attention is directed at whoever is speaking at a given moment, and the defaults in both the UI and how the product functions are optimized for that.

I definitely don't believe that simply taking IRL offline interactions and replicating them in an online container is the way to go.  Similarly, there are technical challenges that avoided by optimizing for a highly ordered, meeting use case.

However, small design tweaks can have a huge impact on how a product feels, and so now that there's a myriad of reasons why people use video conferencing (to socialize, to run exercise groups, to teach and learn, and more) there's a huge opportunity to design products that cater to those experiences and do them well.

I was hopeful that Zoom's APIs would enable one to build more customized apps on top of their infrastructure, but currently that is not the case.

If anything, simply providing meeting hosts with a custom UI tailored to their specific use case would be a huge boost.

There are companies out there that are providing APIs to build WebRTC-based video experiences, and perhaps that is reasonable place to start.  However, there's been a huge gap in performance between Zoom and pretty much any WebRTC-based client that I've tried.  Perhaps that is a function of where the technology is at today, or perhaps that's a limit of the technology itself.