Roblox's future will rhyme with YouTube

Too big to not look vaguely similar...

I often feel the best way to understand Roblox is to think of it as a slightly counter-factual YouTube. Like YouTube (or certainly in its earlier days) it is driven by a young audience. Like YouTube it has spawned a vibrant ecosystem of publishers (MCNs), creators (developers) and content investments (incubators). And like YouTube, it appears destined to become 10x larger than anyone ever thought.

By extension I often think about Roblox’s share price in the form of ‘how would YouTube have performed as a public company?’. An historical trawl of the ups and downs mapped to a notional share price impact suggests that the biggest video platform on the planet would have faced lots of moments when nobody believed that would be the outcome. Today YouTube is worth ~$550B.

Watching this week’s drawdown for Roblox, I feel like it’s another one of those moments. I’ve written about the Lindy Effect of major consumer platforms before and I think that AI puts that into relief for reasons of distribution. It’s really, really, REALLY hard to recreate the kind of consumer reach which spaces like Discord and Roblox have today.

Relatedly it feels like last week’s SaaS/Software sell-off also ignored this distribution reality in enterprise. Byrne Hobart nailed it:

This selloff seems overly fixated on enterprise software companies as software companies, rather than as companies whose core competency is to understand the True Org Chart, figure out a company’s needs, and sell the right products to them.

Can AI natives figure out true enterprise (and SMB) AI adoption muscle and reach before the incumbent software companies (of all sizes) figure out LLMs? Over the next decade (?) I think the outcome will be a mix. I fired up Claude’s Excel plugin over the weekend and immediately loved it (I have historically been a die-hard Gsheets user, resorting to Excel only in times of extreme modelling). Sample size of one but it instantly generated a full-year subscription to MS365 (on top of my Claude Max subscription and token spending). An interesting example of where native AI companies seep into the cracks and crevices of existing software distribution and output positive incremental productivity and revenue.

Other things happening

MTG (where I am on the board) is one of Europe’s least known unicorns. They announced a stellar set of results for Q4 and full year. And excitingly are moving forward with plans to IPO the PlaySimple division in India.

Excluding their IPO plans, there are some growing signals that Discord’s ecosystem is going to get much, much bigger. Levellr (I’m Chairman), which builds enterprise social tools for Discord, continues their mission to have every major video games legend on their cap table announced an investment round that included Mark Pincus (Zynga founder), Bing Gordon (EA), Phil Mansell (Jagex CEO), Frank Gibeau (Zynga CEO) and many others. Wildfire, which connects brands to Discord communities, also raised some investment (from us).

We have quietly started a very specific newsletter for AI Enablement operators (consulting, training, implementation) based on the many conversations that 10xHumans is having in that space. You can kind of think of it as the anti-SaaS newsletter but not in a mean way. Daria has written some very good posts on scaling founder-led sales, org structures and other operational challenges.

I re-read Richard Feynman’s report on the Challenger shuttle disaster. This is a piece of writing that accidentally finds me every few years. It’s short and really worth your time to think about how commercial pressures lead to inaccurate communication of probabilities.