Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Remember that ITP's goal is to break all tracking to begin with, so assuming Apple successfully implements ITP to the degree they clearly intend to, this proposal is built on the ground principles of:

- no tracked link, iframe or not

- no campaign ids

- no conversion POST at all

- no view converstions, no non-last-click attributions

In that scenario, is this proposal an improvement? It sounds like an improvement. It's certainly more than the nothing that anyone will get from Apple's users without it. They didn't have to throw anyone a bone here, and they're going to take massive flak from a few users for doing so at all.

I imagine Apple would be thrilled if advertisers chose the destruction of the tracking-based advertising industry over using this offering. Perhaps advertisers will learn from the music industry's experiences trying to bury their head in the sand re: iTunes and pay attention and participate in designing this system. Perhaps they will not.



There are still ways to do conversion tracking with ITP, mostly moving to first party cookies and I wouldn't be surprised if we see a return of server side tracking like old Urchin so google & FB can store http only secure cookies that are indistinguishable from login/session cookies

If the goal is to stop all tracking why create this (bad) plan in the first place?


The server-side tracking you describe would instantly be corrupted by massive site fraud, as removing the intermediary of the browser from the measurement of “views” would compel people to fake up Apache logs to prove the views they want to be paid for.

The goal is to stop the tracking of individuals by advertisers, without stopping the tracking of ad views by advertisers.


the goal you just mentioned will not be accomplished with what this proposes.

I'm also not as sure as you are that server side ads identity management would be any different than the existing arms race with measurement of measurement meta-ness.

client facing JS is messed with all the time already as well. monkey patching etc. trying to detect monkey patching, toString prototype toStringMetaMeasurement bs etc it will always be a fraud arms race.

to me buying ads is about finding more trust and verifying when possible, obviously any system will be gamed but I'd trust the NYTimes would not mess with FBs server side npm package. and it would make it a lot harder for safari to block when the login session cookie IS the identity cookie FB uses


I mean, pretty much every ad platform has already been pushing for first party cookies due to ITP 2.0.




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

Search: