Unpacking Crowdtangle's breakup with Facebook
My chat with Brandon Silverman, Cofounder & ex-CEO of Crowdtangle
Today’s Scatter Brain is brought to you by AngelList!
AngelList offers startups one platform for cap table management and fundraising. Companies like StabilityAI, Harness Wealth, and Syndicate migrated from other cap table vendors in less than a week to remove the friction from their back office. Learn more by signing up here.
Brandon Silverman’s last day at Facebook was in October 2021. He had been at the Meta since it acquired his startup, CrowdTangle, in 2016. Crowdtangle was the most useful tool to find out what was trending on Facebook and other social networks. Much has been written and speculated about what went down between Brandon’s team and Facebook executives towards the end of 2021 and early 2022.
I talked to Brandon to get his reflections on his decade-long journey and his latest perspective on social media, content moderation, and regulations.
Sar: I have found it strange when third-party players start offering analytics on activity on someone else’s platform or network. When you started in 2011, did you have a lot of paranoia around this existential risk?
Brandon: If you’re an entrepreneur, you’ve signed yourself up to live in a near-constant state of paranoia at some level. But there are ways to manage it and, more than anything, keep it in perspective. There were enough third-party analytics tools built on top of platform data when we started, so we didn’t think it was crazy to get into the space, but we knew we were going to have dependencies right out of the gate, so we did a few things to try and be smart about it.
First, we integrated as many platforms as possible so we weren’t exclusively dependent on one company. Second, we tried to build constructive working relationships with each platform; most of them ended up being our clients. Third, we ensured we were delivering value that the platforms cared about. It was kind of theory of partner transitivity…if we were valuable for x partner and x partner was valuable to platforms, then we would be valuable to platforms. Fourth, we ensured we had exceptional customer service because that was something larger players struggled with, which gave us a chance to differentiate. Fifth, we focused on making products and shipping as quickly as possible.
And, of course, we hoped we could also get lucky along the way.
Sar: How did you think about the state of the broader market?
Brandon: Honestly, we were more worried about existing players in data analytics than the platforms. But it turned out some macro-level dynamics in the industry were trending in our favor, even though we didn’t fully realize it at the time.
In the early 2010s, a lot of the existing data analytic tools approached their work with the notion that they should try and get *all* the data they could and present *all* of it to their users. You ended up with unwieldy, Rude Goldberg dashboards that were incredibly hard to use. It was peak “big data” vibes. When we showed up, there was a lot of white space to go in the other direction to be opinionated about what data mattered, bring a design sensibility to the experience, and above all else, make it easy to use. Fortunately, we managed to get enough right and ended up being a refreshing alternative to the rest of the market.
Sar: Can you talk about an early go-to-market experience?
Brandon: I remember a moment in our first or second year when we had the opportunity to close a $300k-ish annual deal with a client. At the time, our average deal size was around $10k-$15k. So this was the largest deal we would have ever landed, and we were ecstatic. As we got closer to signing the paperwork, I got a sense of cold feet that I couldn’t put my finger on. We were all new to the enterprise space.
We eventually realized that despite how much money we would be leaving on the table, we weren’t ready to get into high-end enterprise deals. We were still iterating too quickly, learning about the market, and deciding what to build. Even a single enterprise client would have meant a level of service & expectations that we could have theoretically provided but would have been disruptive. So we kept the deal, but I negotiated them *down* to a much smaller deal. At the moment, it felt insane, but in the end, our instincts were right.
Sar: What recurring topics did you constantly talk about with your team?
Brandon: Pricing was a constant source of debate among our team, and I’m not sure any of us ever felt like we nailed it. We had a standing monthly meeting where the topic was, “should we completely revisit our entire pricing model?”, lol and we constantly debated every aspect of our pricing (a model based on users or amount of data you used, or how many features you got access to, freemium versus premium, etc.). We also constantly discussed how much capital to raise versus how much to try and bootstrap. In the end, we only raised a small amount, which I’m thankful for, but we were close to raising a lot more money several times.
Sar: What about the CEO’s job was hardest to get comfortable with?
Brandon: I realized early on that your team is the first product you’re building. Your first 0-10 hires are almost existential decisions. But hiring is really hard, and every hire you make comes with some risk. The people who join your team take a big risk themselves…personally, professionally, and financially. As CEO, I took on a lot of emotional responsibility for not wanting to let our team down.
Second is a piece of advice that a friend gave me: the startup experience is an emotional rollercoaster, no matter how well it goes. You’re going to have amazing days with incredible highs where you think nothing can stop you, and (sometimes within the same 24 hours) you’re also going to hit deep lows where you’re utterly convinced the entire thing will fail. The worst part is that that cycle repeats itself for as long as your CEO. You have to keep reminding yourself that the ups & downs are just a part of the process.
Sar: What does Facebook get right with acquisitions? What factors led to the downfall of the relationship?
Brandon: We had a phenomenal experience for the first three and a half years. It was a combination of (1) their ethos in how they had approached M&A, (2) both parties making sure we were aligned on what we each wanted, (3) our team over-delivering on our goals and building up trust internally, and (4) being in the right place at the right time (for a while).
Between 2016, when they acquired us, and 2020, a lot of things changed…some slowly and some quickly. Some of that was the world changing, with Facebook suddenly getting a lot more public scrutiny, the brand reputation taking a huge hit in the US following Cambridge Analytica, and the first signs of regulatory threats popping up. Some of it was our team & what we were passionate about changing as the news industry continued to go through upheaval, and we approached a decade of doing the work. Some of it was shifts in what leadership wanted and their priorities (as well as who was even in leadership). By the fall of 2020, we weren’t as aligned as we had been when we joined (ironically, I think what we were doing was more important than ever to the outside world).
But I think to some degree, that’s pretty normal. If a large platform is acquiring your company, founders should assume that any alignment you have on a mission probably has a half-life of a year or two these days.
Facebook has struggled with what to do with some of its M&As after the honeymoon period was over in recent years. There were unfortunate endings to some of their best acquisitions over the last few years with reasonable and thoughtful founders. I think their reputation as acquirers has declined in the last few years. There aren’t as many founders at the company anymore, and those people can be an important source of innovation…something Facebook has particularly struggled with recently. That being said, the regulatory environment has changed so much it’s not clear how much acquiring they can do at this point, so I’m not sure how much it matters.
Sar: What did you work on at Facebook that you are proud of, didn’t get much airtime, and has now been overshadowed by the following controversy?
Brandon: The vast majority of the work we did never got covered in discussions about CrowdTangle. Our team did so much that I was proud of. Everything from helping protect elections in Brazil, Myanmar, Ethiopia, Sri Lanka, and the US to supporting human rights organizations to empowering fact-checking organizations to work with emergency response agencies in the US and Philippines. We helped teams, from threat investigations to business integrity to global ops, do their jobs better and faster and had a huge impact.
We were also the first dataset available to researchers and academics after Cambridge Analytica, and hundreds of journal articles have been written using our data. We were a critical part of the initial global response to COVID, powering one of Facebook’s first public contributions to the pandemic and working with health agencies worldwide to help better understand what people were saying about the virus. We did a lot of work with underrepresented voices to help bring more diverse publications onto the platform. I spent three years working with lawyers to help launch the first-ever content-based archive of all the networks that were removed for coordinated inauthentic behavior so that academics could study the networks (instead of having them simply disappear from the platform).
Sar: What about FB’s culture that you think is underestimated?
Brandon: When this period of the history of Silicon Valley is written, there will be at least one story about how radical Facebook’s internal culture of openness and transparency was, especially compared to the incredibly siloed & secretive cultures of Snap, Apple, and others. The openness at Facebook made it a fascinating place to work & helped it attract brilliant people. I think it made it a better company by fostering collaboration. But that openness also made it vulnerable and had some material downsides for the company. I worry about Facebook being much more closed off internally, doing less research into their social impact, and, more broadly, other companies looking at what happened and ultimately deciding open cultures aren’t worth the risk.
Sar: The CrowdTangle-Facebook saga was one of the hottest stories last year. CrowdTangle was portrayed as the obvious victim, and Facebook was the undisputed aggressor. What do you think Facebook doesn’t get enough credit for?
Brandon: Honestly, I think Facebook doesn’t get nearly enough credit for simply acquiring, funding, and supporting all the work we did for four years…especially relative to how little transparency efforts were happening at other platforms (Twitter being the exception).
The ending is a different story, but we did a lot of great work for four years, some of which was entirely new and unprecedented in the industry. And I think there’s a sense that our team was like the man who fell off the roof of an office building and shouted to the workers on each floor, “So far, so good!”. And that just wasn’t the case. Most of our work would have never been possible if we hadn’t been acquired and they hadn’t been so supportive for so many years. And in addition to our work, they were the first to start putting out transparency reports on their content moderation and trying to do some really deep collaborations with academics. In a lot of ways, they were genuine leaders in transparency in the industry. It’s part of what made it so fun to be doing our work while we were there, and it’s a reason I’m so worried about them pulling the rug out from all that work.
I also think the discussion about whether or not Facebook is a right-wing echo chamber in the fall of 2020 was never covered well; a lot of that is just Facebook’s fault. I would have loved to see a concerted effort by Facebook to respond in good faith to that critique (despite how frustrating it was to some of them). Do real research into the ideological makeup of the feed, engage outside researchers, release the results publicly, and do the same analysis in different countries. A more introspective response would have put them in a better position to debunk the genuinely unfair parts of the criticism (which there were) and also uncover more true things than some leadership realized (which there also were).
But by this point, there was so much frustration & loss of trust between a few key executives and the mainstream news industry that their response to any sense of unfair criticisms had morphed into a kind of Pavlovian defensiveness at best and a self-defeating lashing out at worst. Combine that with an existential paranoia about being blamed for Trump’s re-election. You end up with a public response that isn’t driven by any strategy but just sheer, unbridled catharsis where all they ended up accomplishing was cutting off their nose to spite their face.
And the irony (or tragedy?), of course, is that during the entire episode, internal ranking teams were working to address some of the issues that were raised, and that work would go on to impact the ranking of political content in the feed at a foundational level.
And maybe one last note: the entire debate was also premised on the notion that you can compare the two media ecosystems in the US as if they were apples to apples, but the truth is that they’re deeply different. Facebook was in a unique position to reveal some of those dynamics and meaningfully contribute to a broader understanding of American media & politics writ large. But instead, they chose just to try and make themselves feel better. It was such a missed opportunity."
Sar: What’s missing or not appropriately debated in the public discourse around content moderation and platform transparency?
Brandon: I think this is starting to break through in the overall discussion, but the misinformation and disinformation discourse over the last several years frequently missed the forest for the trees when evaluating how healthy an information ecosystem is. False and intentionally misleading information is only a very small fraction of the ecosystem. For instance, there is a lot of content that is technically accurate and non-violating but can be deeply misleading. There is way more violating content from mainstream actors, including politicians and elected officials, compared to troll farms or bots.
I also think there’s a growing recognition (Elon Musk notwithstanding) that there’s no easy way to manage global public spaces dedicated to free speech. Setting rules is insanely messy, full of hard tradeoffs between competing values with unintended consequences, not to mention competing laws around the world…multiplied by different communities & local nuances…and that’s to say nothing of executing those things once you plant all your flags. It’s a set of insanely complicated scientific & mathematical problems sitting on top of a more complicated & messy mass of humanity.
There’s also no shortage of ways to screw it up, too many of which we’ve seen in the past and too many that we’re still seeing today. But that’s why I’m excited about the growing attention paid to topics that go beyond content moderation, including the basic design affordances of platforms, the ownership & governance models, what the values of a platform are, how well the values are reflected in its impact and the role of transparency in helping govern & holding them accountable.
Sar: Knowing how everything played out, do you have any regrets?
Brandon: There are things I know I could have done better along the way, some of which I’ve thought a lot about at different points, but I mostly feel grateful about the entire experience. I feel grateful that our startup didn’t completely fail, which is what happens to most startups. I feel grateful we had an exit that meant a lot to our team and gave us a home to continue & grow our work for years. I feel grateful that we put together a special team that still supports each other, especially during downswings in the economy. Most importantly, I feel grateful we were able to help so many organizations & people around the world doing important work.
That said, I worry a lot about the organizations that rely on this data to do their work. I regret what might happen if Facebook continues to move away from transparency or if they only share data with a small group of researchers instead of working with the rest of civil society. There have already been a lot of moments over the last year when I wished our team could have been around to help partners I know were doing important work, and that was hard (including around the US midterms, the war in Ukraine, the election in Brazil, and more).
I’ve gotten involved in the regulatory work around transparency and trying to move to a place where this sort of data sharing and collaboration is legally required and not subject to the whims of a few executives or the predictable dysfunction of large corporations.
Sar: What’s capturing your attention these days?
Brandon: Besides working on the coming regulations of social media around the world, I’ve been advising a handful of new groups trying to imagine what the internet & social media could look like if we were more intentional about building it to support democracy. One of my favorite organizations is New_Public which is doing cool and timely work and where I’m on the Product Advisory Council.
I’ve also been trying to catch up on the progress in clean energy & decarbonization (we’re about to get our first heat pump at our home!), as well AI. As a creative exercise, I’ve been trying to learn Unreal Engine 5, which is mind-blowing and super fun to play with (I’m trying to create a mini-game based on the book Piranesi by Susannah Clarke).
Sar: What are we not paying enough attention to these days?
Brandon: I wrote a memo on Facebook about the role that fear plays on social media & the internet, and it’s something I still think about. I’m reading Martha Nussbaum’s “Monarchy of Fear,” along with some newer books about the popularity of authoritarians around the world and how they use fear, including “Popular Dictatorships” by Aleksandar Matovski. I want to explore the idea space a lot more, and I think it might be one of the most important underlying dynamics that we’re not paying enough attention to.
Previous chat :
Future of coding in India and on mobile with Amjad Masad, Founder of Replit