We track for targeting, not for attribution. I don't see any benefit for the targeted individual or the advertiser coming out of this proposal. This proposal only brings benefits for the parties on the publisher chain who now have one more angle for committing fraud.
I am aware of ITP. There is no ad-tech without targeting. There are >50 million ad opportunities out there at any second, you need to target to delivery the message to a proper recipient. Attribution is a bottom of the chain problem, often problematic due to fraud.
I don't think people care that Amazon (for example) suggests things to them based on previous purchases because they have a first party relationship with Amazon. Likewise, when I search for "honda" I'm okay with Google showing related ads to the right of my search results. Targeting isn't bad.
The concern is when user data is shared with third parties.
Of course we do that. But then how do you validate that wasn’t a bot generating a request trying to increase the publishers traffic? Do you trust the publisher blindly? You need some sort of insight into the web request other than a single integer counter claiming an impression occurred in order to validate the individual.
That's not how the business model works. You pay for the impressions delivered and the cost of the data used for targeting. If I get no conversions, who pays Slate.com for the million of impressions that nobody clicked on?
