Hacker News new | ask | show | jobs
by niknetniko 1255 days ago
Am I missing something here? I understand the "The IAB loves tracking users." part of the title, but where is the part about "But it hates users tracking them."? Based on the title, I would have expected an exception in their standard for themselves, or a story about how their own user tracking technology was used against them.

However, there is none of that. Just that the IAB does not want to make it easier for users to escape their tracking (which, given their purpose, is unfortunately entirely expected). What justifies "But it hates users tracking them."?

7 comments

The later part refers to the plus addressing, like "han.solo+github@gmail.com", which people use to so that "If they later start receiving spam to that address, they know the service has leaked or sold their info.". Now the IAB requests that advertisers should normalise such addresses by dropping the part after the plus sign, and therefore effectively stopping users from "tracking" the advertisers.
This is dangerous as some mail servers could consider plus addresses unique.
(Web) developers have gotten lazier and simply don't care anymore. The fact of the matter is that if you don't host your email with one of the big three, some services are probably not working anyway. I'd also like websites to show me things like news and recipes without having them run javascript, but apparently this means I deserve white pages because writing HTML is too much of a hassle for the modern web developer.

It's quite sad to see. It's also the reason I'm using somethingunique@domain.tld;, if cyberstalkers start normalising to a domain, they'll only hurt their own business.

> I'd also like websites to show me things like news and recipes without having them run javascript, but apparently this means I deserve white pages because writing HTML is too much of a hassle for the modern web developer.

Often enough the content is there but just hidden until the JS loads for ... reasons, idk.

nobody cares, email, along the majority of the internet as "computers talking protocols", is dead. The absolute majority of email is handled by gmail or microsoft, accounting for >80% of MX servers in the wild by the data i had five years back. I'd imagine the share of the duopoly is even larger today, considering how difficult it is to get into inbox these days.
EMail is more alive than any other federated communication protocol on the internet. It is only rivaled by phones and physical mail.

Defeatism here helps noone.

In the linked document they say to only do it with @gmail.com addresses.
>and therefore effectively stopping users from "tracking" the advertisers.

I guess that's the nefarious explanation, but there's a more benign one: if you want to correlate user behavior, you need some sort of normalization, otherwise john.doe+apple@example.com and john.doe+amazon@example.com would show up as different "people" and cause match rates to suffer. Sure, getting tracked isn't great, but it's not exactly the hypocrisy rage-bait that the OP is implying.

Why would a person want advertisers to correlate their behavior? What’s in it for them?
In a logical world it would mean that you see fewer ads. If your matches increase, your price per ad goes up so the service needs to show you fewer ads to hit their ad revenue target for you.

But if you think that will happen I have an East River transportation startup in New York that is seeking an angel investor.

Revenue targets are certainly dynamic but avoiding user churn is important too.

So you want to keep cost per impression up. You would not want to saturate and devalue.

Better to play 10 ads at 10c each vs 20 ads at 4 or 5c, as high ad load impacts users propensity to return to the service.

> the service needs to show you fewer ads to hit their ad revenue target for you.

In a logical world, yes.

In a capitalist world, that revenue target goes up every year. Apple became the richest company on earth selling hardware, yet here they are now drowning their software with ads.

We really should just cut out the middle man here and ban ads entirely. If an ad broker wants to pay me to watch an ad, pay me directly.
> drowning their software with ads

Where? I use Apple hardware basically exclusively. Are they that good in hiding the ads, or are you exaggerating a bit?

Ostensibly it's so that the user can get more relevant ads.

In practice, it's not of course, but that's the answer you'd get if you ask them.

Of course that's the goal. The IAB isn't an NSA front. The problem with advertising is not the primary goal, but all the secondary things that can happen.
lokedhs said "ostensibly" because the internet advertising industry has long maintained the pretence that tracking users and showing them relevant ads is helpful to the user, when the truth is the advertising industry cares about the opinion of users like the thanksgiving industry cares about the opinion of turkeys - which is to say, not at all.

There's a reason these things are opt-out rather than opt-in.

The answer to marcus0x62's question - why would a person want advertisers to correlate their behavior - is that they wouldn't, and if they want to advocate for their own self-interest they should install an ad blocker.

If it's the goal, you'd expect ads to be actually be more relevant and useful to users.

They don't, which has been shown in studies. What has been shown is that showing the same things people already bought give people regret which increases total amount of purchases.

In other word, the goal is to not to give users a good experiences watching ad's. It is to make them buy more, which is an orthogonal goal.

Let’s say you’re DHS. You contact some person and have a conversation like this

Govt: “I need IP addresses, ideally cellular and known public wifi, of a person using this email address”.

Data broker: “Here’s the list including the most recent cellular IP address associated with that person at this timestamp and their most used public wifi locations.”

Govt: “Hey, cellular provider, where is this subscriber right now?”

Provider: “Here’s the lat/long, last seen 1 second ago. Happy hunting!”

I disagree: The main problem with advertising is with its primary goal, which is manipulting people into excess consumption. There are of course other secondary issues as well, but even without them ads are already a net negative for society.
Yeah; that's not benign. I don't want my behavior being "correlated" by a shady group of companies whose sole purpose is learning how to better manipulate me for their own profit.
I never claimed as such. From the comment you replied to:

>Sure, getting tracked isn't great, but it's not exactly the hypocrisy rage-bait that the OP is implying.

> if you want to correlate user behavior

That in and of itself is nefarious.

I suppose we should start using aliases then instead.

Ultimately they cannot win this fight.

They can’t win this fight against people like us, but for the rest of people it’s a mess.
>> What justifies "But it hates users tracking them."?

Users can tell which site gave away their email address by using the variants discussed. It's not tracking in the same sense, but it does allow tracking who respects privacy. It also allows throwing away junk mail where someone required an email address just to (for example) make a sale or use their wifi.

It was a quicker way of me writing "but hates users being able to track what their members are doing with user data".

The reason my email address for this site is `+ycombinator@...` is that I can track when dang decides to go rogue and sell my email to a Nigerian Prince.

Think of it like sousveillance.

Have you ever received any spam addressed to that account? I've signed up for countless services over the past decade, the only ones to ever spam me were university consulting groups back from the time I was applying to grad school.
Yes. I frequently find services which - either maliciously or through incompetence - allow my email to be used for 3rd party spam.

Even some big companies aren't immune to a dodgy contractor walking off with a contact list.

Let’s say you use a “plus” formatted email, I.e foo+github@gmail.com

Then when you receive spam/unsolicited marketing emails, you can see to which email the spam was sent, and therefore which company sold your data.

This suggests the only way to keep this behaviour is to have your own email hosted and use a truly different email per service.

They don't want us to be able to see which websites are creeps and collaborators building secret dossiers on us.
> What justifies "But it hates users tracking them."?

Nothing, I think, since there’s no indication the normalized email will be used to send email.

The normalization is for connecting together identities against the wish of the user, which is a different issue.

I suspect that part of the title is referring to the section that discusses using different email addresses for different services, so you can tell who sold your email address.