What Open Source Taught Me About Coalitions

Written in 2021, about work done in 2020. The pandemic context has receded; what happened to those teams hasn’t.

While pandemic anxiety raged, the global tech community collaborated in a way that would have lasting impact. I was fortunate to be one of them, and here are some of the lessons I took from it.

Lesson 1: Open source works

I was at home in NYC in March of 2020, listening to the city’s eerie silence and wondering how I could be most useful. I came across an appeal to join “Wir vs Virus” — “We vs. Virus” — a hackathon sponsored by the German government, and was intrigued. I soon found myself on Slack with about 28,000 similarly motivated people, brainstorming how technology could address pandemic challenges: telehealth, e-learning, getting food to the elderly in a lockdown.

I advised a few teams focused on what were then called contact tracing apps, later exposure notification. For those to work at all, the apps had to be able to end up on everyone’s phone — which meant inventing a protocol, a common language letting devices talk to each other with global interoperability. It also had to not infringe on its users’ privacy, a top concern especially in Europe. You don’t want to inadvertently build a mass surveillance tool.

Our work was not among the awarded hackathon winners. That didn’t mean it wasn’t valuable: it went on to become one of the foundational building blocks of today’s global exposure notification apps.

Coalition building

About a week after the event, I went back into the Slack channels and checked in with the teams. There was a clear sense that the work was useful, but no clarity on where to take it or how to operationalize it. If our ideas were going to help, we needed to speak with one voice. So I brought the teams together to join forces and merge their work, to stop duplicating effort.

I spoke with colleagues tackling similar projects in the US, and together with Harper Reed — an entrepreneur who had served as CTO for Barack Obama’s 2012 re-election campaign — and Jenny Wanger, a tech leader with a product management background, we decided a global merger of these efforts was necessary. We made it happen with a few long Zoom calls, many individual phone calls, and a Slack channel where we invited everyone we knew. The enthusiastic collaboration of hundreds of people across the globe was beautiful to watch.

Lesson 2: Open source’s problems yield to a uniting cause

The familiar complaint about open-source work is that it takes forever, or that it’s full of highly opinionated people who block everything. These can be real hurdles — for open source or for any committee-led project. But they can be overcome with good leadership, friendly guidance, helpful moderation, and a clearly articulated common goal. Here it helped enormously that everyone was pointed the same way: safely reopen the economy, safely reopen travel, save lives.

As we converged, we merged a German-made protocol called STRICT with one developed in the US called CEN, naming the result the Temporary Contact Numbers Protocol, or TCN Protocol. It was an audited, open-source, decentralized, anonymous exposure alert protocol in which the central reporting server never has access to content logs. As apps were built around the world, the protocol was what would let them talk to one another.

Coordinated work needed a home, so the TCN Coalition was founded on April 5, 2020, as the global umbrella organization. In the end, our protocol was one of the two adopted by Apple and Google and developed into what is now commonly called Google/Apple Exposure Notifications, or GAEN.

The coalition kept going as a global community building privacy-first digital tools for public health. In July 2020, TCN Coalition merged into Linux Foundation Public Health, with Jenny leading the Implementer’s Forum — coordinating implementation teams around the world and publicly documenting best practices.

Lesson 3: Team building is everything

I’m proud of this work and honored to have been part of it, in a leading capacity at that. But a key reason we could do it was the caliber of people involved.

I won’t list endless thank-yous, but some people deserve acknowledgement: Tina White, James Petrie, and Zsombor Szabo, who founded Covid Watch. Dana Lewis and Scott Leibrand, who founded CoEpi. Martin Häcker, Kilian Holzapfel, and numerous others behind ito. Bruno Bowden, who led the effort for the official WHO app. Peter Giese and our other early supporters at SAP. Ellie Daw, T.E. Shaw, Henry de Valence, and countless others who volunteered their time.

And it isn’t only about smarts. It’s about grit. Two collaborators who stand out are Nele Quast, a PhD scholar from Berlin, and Clara Fischer, a startup scout and communications expert from Leipzig. They were on the German team behind the ito app. On incredibly short notice — I think it may have been under twelve hours — they were invited to pitch to the Dutch Ministry of Health, Welfare and Sport in The Hague in April 2020. Driving across Europe, Nele and Clara got past closed borders, shuttered hotels, and no sleep to present alongside experts from Accenture and CapGemini. And they nailed it.

The Netherlands ultimately chose none of the apps, but Clara’s and Nele’s presentation during that multi-day, televised selection process convinced the citizens: ito was still the favorite when it went to a public vote.

Their willingness to go the extra mile — along with the unseen background work of dozens of other member teams supporting them — was a reminder of what you can move when you surround yourself with the best people.

What transfers

The pandemic framing has aged out. The mechanism hasn’t. A scattered group of smart people with overlapping work and no shared language will stay scattered until someone does the unglamorous work of naming the thing they’re collectively building and getting them to agree to it.

That’s the same work, structurally, that I now do with clients: finding the one articulation a group of people can all stand behind, and making it specific enough that they can act on it without checking back in. It’s what a Point of View document is for. In 2020 it happened to be a protocol.