I dunno if the employees were ever really needed for scale. WhatsApp famously had 300M users and 13 employees at the time of acquisition; Instagram was something like 50M users and 55 employees. If you know what you're doing software scales basically infinitely, and the employees are there to make the software just slightly more tailored to specific user populations (and because upward career mobility for managers involves having more headcount). Yeah, building a revenue model takes people, but Valve employs only about 400 people and makes billions, as do various quant hedge funds like DE Shaw or RenTech.

The insta/whatsapp/plentyoffish model works if you're very lucky with both product-market fit and the technical constraints of the product itself. If you have something that technically scales extremely cleanly, it basically sells itself, and it doesn't need feature churn to retain or gain users, you're golden. I do think more businesses could do with checking whether they do in fact have that lottery ticket before hitting the scale button; there aren't that many examples around.

> Valve

Arguably a monopoly. They've got a product that sells itself with very low infra overheads for the income.

> Hedge funds

Very different model. I don't think the same intuitions apply.