That's not the case. WhatsApp barely holds any data. The user doesn't have to give over their personal details. It ties them to their mobile number & messages are stored only until they're pushed to the client. Once pushed they are then removed from WhatsApp servers.

I'm curious: how do you know that they're removed?

They pinky sweared, of course!

/s

If you map some 5-10% of these phone numbers to real users (which facebook can do), then you can map almost all of the remaining phone numbers to their real names as well, using the graph info.

There were some recent papers on algorithms how to do it, and they managed to match ~90% of 'anonymous phone numbers based on the facebook graph data. You stay anonymous only if all your friends and their friends care about anonymity as well.

even if removed, WhatsApp could store metadata like who is talking to whom. This is valuable to try to identify the trendsetters and key people to market to. With a bit of text analysis like google do with gmail, they can personalise ads more. all very very valuable data.

You underestimate the value of the graph data.

Great post, was laughing throughout.