Roblox and the no-win continuum of kids digital design
Last week Hindenburg Research announced their short position in Roblox. This was ostensibly based on claims of usage inflation with a mostly ad hominem attack on their child safety policies thrown in for good measure. In general, I quite like what Hindenburg does in the world but I was disappointed in the weakness of this report1, a view seemingly shared by the market (Roblox’s share price has since mostly recovered). I’m not going to write about the report itself because I don’t really have much to add beyond what Joost has already written.
I did however, want to talk about the challenges of building digital consumer services which support kids. There is a tendency for people to criticise this area with a simplicity afforded to those who have not had to actually tackle some of these problems. In fact, even the word ‘problem’ is a misnomer: in the kids’ space you are not really solving anything, you are choosing points on a continuum between four fundamental sets of design trade-offs.
Features v control
Although these trade-offs are not in any order, I think this one is logically first. The more you deprecate features for kids to try and deliver a completely safe experience (e.g. social), the more they will ignore the ‘kids’ option and just go straight to an adult version if they can. Kids do not care about your app store labels.
Privacy v safety
The more privacy you want, the less safety you get. A good example is age-gates, one of the best lying incentives our civilisation has ever devised. You can get step-change improvements on this with face-scanning or, which I am fond of talking about and someone needs to build, a trusted age API. But to some degree (although not as much as Taylor Lorenz claims) you give up some aspects of privacy.
Non-commercial v sustainability
All things being equal, a parent would probably prefer their kid plays a game rather than participates in an economy. But presumably that parent would also prefer a professional trust and safety team behind the scenes also. An extension of this debate is paid versus free (ads).
Parenting v technology
Look, parenting seems extremely hard in 2024. And you would probably not be shocked to know the number of parents who ignore meticulously developed parental gaming tools and let kids use their account because it’s just quicker and dammit I’m tired. Ultimately technology can’t parent and parents can’t technology. Humans are tricky.
I have a teeny tiny suspicion that all of this might have originated from one Hindenburg analyst’s parenting experience. We’ll presumably never know but it is not the first time I’ve seen enormous amounts of capital put into play on the basis of parents reacting to their children’s behavior.