❓ Problem

Kids have always played video games. The difference now is that the games are social by default, and the other players are strangers. Roblox alone runs at roughly 132 million daily active users, and among users who have completed an age check, about 35% are under 13. That is tens of millions of children logging into a live social space every single day. Oklahoma's Attorney General cited the company's own figure that two-thirds of US children aged 9 to 12 are active on the platform.

Parents know this and mostly shrug, because the alternative is telling a nine-year-old they cannot play the only thing their entire class plays. Norton's 2026 Connected Kids research found that among US parents whose child plays chat-enabled games, 64% say their child has chatted with a stranger, 36% say their child has received inappropriate messages, and 38% say their child has been encouraged to leave the game and continue the conversation somewhere else. That last number is the one that should stop you. The game is not usually where the harm happens. The game is where the contact happens.

Here is the kicker. The industry response has been to bolt safety onto a product that was never designed for it. Roblox now runs mandatory age checks for chat, has settled with three state attorneys general for $35.8 million, and is defending a federal MDL with more than 160 pending suits plus separate actions from Texas, Tennessee, Iowa, Florida and Nebraska. Every one of those patches makes the product slightly worse for the child and only slightly safer for the parent, because the underlying design assumption is unchanged: anyone can play with anyone.

Meanwhile the only two options a parent actually has are both bad. Turn multiplayer off and you have removed the entire point of the game. Leave it on and you are trusting a moderation system that regulators in five states are currently suing over. Nobody has built the obvious third option, which is to keep the multiplayer and change who is in it.

✅ Solution

A private gaming network for children, where the multiplayer graph is built entirely from families the parents already know.

• Original games in the formats kids already love. Obstacle courses, building worlds, pet sims, tycoon loops. Not clones, but the same core pleasures, built as a small first-party catalogue rather than an open user-generated marketplace. You control every asset in the product, which is the whole point.
• The friend list is a parent list. A child cannot add a player. A parent approves another family, and the two households are connected. Invitations flow adult to adult, which means every single connection in the network has a real-world relationship behind it.
• AI characters and recorded playthroughs solve the empty lobby. The killer problem for any small network is that nobody is online at 4pm on a Tuesday. Clearly labelled AI companions and recorded ghost runs from other players keep worlds populated and competitive without inventing a stranger. This is exactly what indie developer Adam Lyttle prototyped for his own son, and it is the mechanic that makes a closed network viable at small scale.
• One family subscription, nothing else. No advertising, no loot boxes, no in-game currency, no purchases. The parent pays once a month and the entire monetisation surface a child touches is zero. That is not a feature list, it is the pitch.

The wedge is deliberately narrow: one excellent game, one school year group, one set of parents who already talk to each other. The asset underneath is the parent-verified household graph, which is the part nobody else can copy quickly.

📊 Key Numbers

Market size

• There are roughly 33.6 million US families with children under 18. Around half have a child in the 5 to 12 bracket where this product lands, so call it 17 million US households.
• Add the UK, Ireland, Canada, Australia and Western Europe and the addressable base is somewhere near 45 to 50 million households. At a $120 per year family subscription, that is a TAM of roughly $5.5 to $6 billion.
• SAM: the families who both have a child already inside a chat-enabled social game and a parent worried enough to act. Norton's data puts 64% of parents in the first bucket. Take a conservative 40% of the TAM and you have ~19 million households, or a serviceable market around $2.2 billion.
• SOM: 0.5% to 1% of that within five years is 95,000 to 190,000 paying families, or $11m to $23m ARR.