On Platform Shifts and AI

At TCV’s Engage Summit in 2022, I gave a take on finding your next wave of growth, which you can read here. Sam Shank, founder and CEO of HotelTonight, asked me the question that everyone asks growth people, “What new channels are you seeing new consumer companies take advantage of?” My answer was disappointing as it always is, “What new consumer companies? Discord is the last one I have seen grow organically for a long period of time. Tiktok spent many billions of dollars buying up every ad they could on Facebook, Snap, and Google properties. It’s not really replicable.” I then proceeded to explain that consumer companies tend to arrive in droves during platform shifts, and we haven’t had one since mobile. But AI could be coming (editor’s note: it did). Sam quickly pointed out that while AI is a potential technological platform shift, it is not a distribution platform shift. And it’s distribution platform shifts that create new consumer opportunities. I’ve thought about this conversation a lot, and think I have a better framework to both describe what Sam was describing, and what that means now that the technological platform shift clearly arrived when ChatGPT came out.

What separates a major platform shift from a minor platform shift is a platform shift that enables both a technological shift (new ways of making things possible) paired with a distribution shift (new ways of reaching people with it). The internet and mobile both created new technological and distribution shifts that enabled lots of new multi-billion dollar businesses to be created, whereas “cloud” as an example made new things possible without any new distribution (favors b2b innovation) and crypto arguably enabled new forms of distribution (tokens), but didn’t fundamentally make many new things possible with technology. So nothing can stand the test of time in that space. I’d argue the only companies that have found product/market fit in crypto are companies that either enable or catch grifters. A more pithy way of saying this is the crypto space has created more criminal convictions than companies with product/market fit. Other things VCs have hyped in the past as potential platform shifts have largely neither made interesting new things possible nor driven new distribution opportunities (NFC, VR, Internet of Things, et al.).

What I realized having gone through the internet and mobile platform shifts is that the technological and distribution shifts did not happen at the same time. Platform shifts that create both technological and distribution opportunities happen in a sequence, not all at once. The internet created websites, but the search engine wouldn’t come along until later to become the dominant form of distribution. Mobile created mobile apps, but it was Facebook mobile ads, not the App Store, that became the dominant form of distribution for mobile apps. So, AI has come out and definitely created a technological shift that enables new ways to solve problems that couldn’t be done before. But AI lacks a new distribution channel. ChatGPT is “not it”, as the kids would say. At least not yet.

So, today, that means the traditional distribution methods need to carry the distribution of AI innovation. This favors either: 

  • incumbents who already have distribution
  • startups that can leverage traditional channels such as sales, virality, user generated content, or paid acquisition because their product value is deeply innovative and very marketable

But, this may not be forever. As I mentioned before, we shouldn’t really expect new distribution shifts to have happened yet. The App Store launched in 2008, and even though there was fervor around discovering apps on the App Store for a while with the “there’s an app for that” campaigns, that fervor died as did most of the apps featured. It was when Facebook launched mobile ads four years later in 2012 that apps exploded into multi-billion dollar companies. This is similar to the internet. People started getting online around 1994. Google didn’t come out until 1998. Sure, there were search engines before that (Lycos, Yahoo!), but they lacked the predictable distribution of Google. Word of mouth can’t scale technological shifts alone. They need scalable distribution methods, and usually new ones that take time to become obvious.

So, as an operator, this feels like 1997 or 2008. The Google and Facebook mobile ads of AI haven’t come out yet. Most of the companies that exist will die in the next five years like the internet bubble as they lack sustainable business models and distribution, but there are a few that won’t (Amazon, Ebay, OpenTable et al. survived the internet bubble), and much of the next gen after this wave will become very large. And unlike the internet bubble, incumbents are on top of it and many will do quite well capitalizing on this shift. Some will get destroyed, of course.

We also can’t bet on a new distribution channel coming for AI though. With every generation, companies that reach massive scale have gotten more efficient at preventing other companies from growing on top of them, at least for free. Google created a scalable way for companies to grow both organically with SEO and by paying for it with Adwords, and it still works decades later. Facebook, after flirting with a similar strategy to Google, decided to charge companies for all distribution on its platform. So, if this distribution channel never materializes, expect the impact of AI at consumer scale to be mostly coming from consumer companies that already have consumer scale vs. a bunch of new Facebooks and Googles. I’m rooting for those new distribution angles myself though.

Currently listening to my Future Bass playlist.