Friend scraping expanded 270,000 paid participants into data on 87 million people

Evidence

Facebook estimated that Kogan’s data set likely contained information on up to 87 million people worldwide (270,000 multiplied by an average of about 322 friends per user). In the UK, a couple of thousand initial participants ultimately yielded data on about 1 million UK residents. The 270,000 paid participants returned data on up to 70 million Americans specifically – 30 million of which were shared with Cambridge Analytica, who paid Kogan about £800,000 to get the data. — Sander van der Linden, Foolproof, 496

Categorization

• SEED: The 87-million-from-270-thousand multiplier is a vivid concrete figure for an essay on how consent mechanics fail against network effects.
• XREF: Parallels Milgrom’s finding that relationship networks substitute for direct individual knowledge, but here weaponized for surveillance. — connects to: Seed participants’ Facebook friends expanded 270,000 people into 70 million profiles — Sander van der Linden, Foolproof: Why Misinformation Infects…, 496; Facebook’s friend-data scraping multiplied a small consented sample into millions of profiles — Sander van der Linden, Foolproof: Why Misinformation Infects…, 495; Payments attracted the 270,000 quiz takers whose data seeded the dataset — Sander van der Linden, Foolproof: Why Misinformation Infects…, 494

Broader Connections

• Seed participants’ Facebook friends expanded 270,000 people into 70 million profiles #corroborates (0.95) — Sander van der Linden, Foolproof, 496
• Facebook’s friend-data scraping multiplied a small consented sample into millions of profiles #corroborates (0.90) — Sander van der Linden, Foolproof, 495
• Payments attracted the 270,000 quiz takers whose data seeded the dataset #expands (0.80) — Sander van der Linden, Foolproof, 494