Hacker News new | past | comments | ask | show | jobs | submit login

Select * from fb_users join wp_users on wp_users.id = fb_users.phone_number

A really basic query that will give you all the fb users who have listed their phone number.

Also i just realized while typing this, i think fb started allowing logging in using your phone number around the time they bought whatsapp. So even the profiles that didn't give their phone number to fb, they could get it via this logging system. ( I guess i was a victim of that)




I think more than half of the FB users also use the app or facebook messenger, which gives them access to their phone number through app permissons alone. Additionally they have access to the contact lists of both apps and can correlate names and phone numbers from there. Even if someone doesn't use the FB app and hasn't listed his phone number in FB - if any friend has him as contact in FB and Whatsapp it's quite trivial to match the IDs.


Exactly what I had in mind.

You could also complement it with phone numbers of people with 2FA enabled.


FB does not require phone numbers so you are just decorating with data that most users do not provide to FB. Not the same thing.

Not saying FB could not join the tables given how much data they have, but it is not nearly as simple as this.


And that's why i said users who have listed their phone number.

They can get phone number from users who use their phone number to log in but not have listed it in their profile (me), people who have enabled 2FA or from messenger. It's just a simple sql join. Ofcourse not at FB scale but this is me thinking about this problem for 10 seconds and also i'm in probably not even as intelligent as the dumbest guy working at fb.


And then there are people who just use a Google Voice number on Whatsapp instead of a real phone number, and many don't even have a phone number listed on Facebook. I'm actually surprised that anybody would give their real phone number to Facebook, Whatsapp, or any of the other "social networks"...


> I'm actually surprised that anybody would give their real phone number to Facebook, Whatsapp, or any of the other "social networks"

The only use for my phone number is registering for stuff like WhatsApp. Why would I want one if not for stuff like that?

(note: talking about the number itself, not a smartphone. I use a smartphone for lots of other stuff)


But you can give Whatsapp a Google Voice number instead of your real phone number. That way if they decide to sell it as part of your profile, you can easily change/cycle it (or apply better filtering, DND, etc.)


AFAIK, Google Voice is only available in the US, and this trail is in the EU.


2FA + the very fact that many - myself included - use Facebook, et al. primarily on a smartphone.


>with data that most users do not provide to FB.

And you're basing this on... what? Facebook nags you EVERY TIME you login for your phone number. EVERY TIME. I can tell you every last one of my friends who is non-technical has put in their phone number simply to get rid of the nag message. While my evidence is anecdotal, it flies in the face of yours. So I assume you have a citation for your "most users" comment.


So i did some search. Facebook app is installed 1 billion - 5 billion times on android alone. Facebook app has access to your contact card, your contacts and bunch of other stuff. Now let's say there are 20% of people who review these permissions and don't give fb access to everything. Even with that it leaves ~800 million apps with access to phone numbers. And this is Android alone.

I would be pretty surprised if evgen had same views if he wasn't working at fb.


Have left and have been working on my own startup for several months now, but I am aware of how some of the data is siloed at FB and also that WhatsApp was run very independently (as in, almost no operational contact between the two companies) up until the summer at least. I don't doubt that FB can tie a specific FB user id to a WA user id, and if they tried hard they could tie a large fraction together using a variety of data sources (certainly not a 'one-line SQL query') but between 10-25% would fail due to bad data on either the FB or WA end. A partially complete data set is not as useful as you might think...




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: