It turns out that as Jeff Bezos's empire expanded to deliver us everything from dental floss to existential dread within 48 hours, the number of people attempting to steal our passwords grew in almost perfect synchronisation, as if the universe had decided that economic growth and criminal ingenuity were merely two expressions of the same underlying principle. One wonders whether the criminals were simply following Amazon's lead, or whether they were both independently discovering the same terrible secret about the internet. Either way, a 96% correlation suggests something has gone quite wrong with how we've arranged things.
The more likely culprit is that both trends simply rode the wave of global internet adoption and digital commerce expansion between 2005 and 2022 — a period when the online economy went from niche curiosity to the central nervous system of commerce, and when roughly four billion new people gained internet access. As more people bought things online and more criminals discovered that phishing required only a computer and the kind of patience that doesn't require food or sleep, both datasets climbed together like two vines on a fence, each following the sun but oblivious to the other. To put it in scale: we went from roughly 250 million internet users in 2005 to nearly 5 billion by 2022, and in that same period, phishing attacks grew from thousands to millions annually.
The correlation tells us almost nothing, which is precisely what makes it so interesting — it's a reminder that the internet itself is the variable that explains both Amazon's ascent and the flourishing of increasingly sophisticated fraud. We see patterns because patterns are everywhere; the real question is whether any of them are trying to tell us something useful. The datasets move together, but they're holding different weapons.
As an Amazon Associate, getspurious.com earns from qualifying purchases. Learn more.
Want to learn more about why correlations like “Amazon annual revenue” vs “Phishing attacks reported annually (worldwide)” don't prove causation? Read our guide to statistical thinking.