WICG / floc

This proposal has been replaced by the Topics API.
https://github.com/patcg-individual-drafts/topics
Other
936 stars 90 forks source link

How can we ensure the open web is not destroyed? #109

Open lukewhitmore opened 3 years ago

lukewhitmore commented 3 years ago

I've been thinking a lot about this proposal, and I've been trying to understand the potential motivation of Google and how the situation could potentially play out to provide Google with an advantage.

dmarti commented 3 years ago

Some other mainstream browser vendors have not announced implementation plans, but they have not committed to not doing FLoC either. (Source)

Web advertising is generally placed through a real-time bidding process, so if legit advertisers begin using FLoC extensively, then users of non-FLoC browsers will not get bid on by the higher-bidding FLoC-using campaigns and end up receiving more lower-quality or deceptive ads. Non-Chrome browsers might end up being able to give their users a better total experience by implementing the FLoC API (with additional protections, if needed) than by leaving all their users in the "null cohort".

michaelkleber commented 3 years ago

Hello @lukewhitmore,

FLoC is one of several Chrome proposals with the common goal of letting the advertising-supported open web continue to function in a world without cross-site tracking. Other major browsers agree that this is a worthwhile goal, and the W3C is home to an ongoing discussion of the best way to accomplish it. If you wish to join in on that conversation, I would suggest participating in the W3C Improving Web Advertising Business Group.

But if your goal is to talk about one browser causing self-preferencing lock-in by behaving differently from the consensus of other web browsers, then you are aiming in the wrong direction. Please take a look at https://infrequently.org/2021/04/progress-delayed/ for a better understanding of the state of cross-browser compatibility and web standards.

lukewhitmore commented 3 years ago

Thx @dmarti

Web advertising is generally placed through a real-time bidding process, so if legit advertisers begin using FLoC extensively, then users of non-FLoC browsers will not get bid on by the higher-bidding FLoC-using campaigns and end up receiving more lower-quality or deceptive ads.

Why would it stand to reason that users would receive lower-quality or deceptive ads if they used a browser that didn't implement FLoC?

lukewhitmore commented 3 years ago

Thx @michaelkleber

But if your goal is to talk about one browser causing self-preferencing lock-in by behaving differently from the consensus of other web browsers, then you are aiming in the wrong direction. Please take a look at https://infrequently.org/2021/04/progress-delayed/ for a better understanding of the state of cross-browser compatibility and web standards.

My goal is to understand the landscape if FLoC is implemented by Chrome, and not by other browsers; and ultimately how advertisers are likely to react in this scenario.

michaelkleber commented 3 years ago

In that case, note that third-party cookies are still available in Chrome today, and have been removed in other browsers. Many publishers and advertisers already have strong opinions about the impact this has had on monetization of the open web.

lukewhitmore commented 3 years ago

@michaelkleber

In that case, note that third-party cookies are still available in Chrome today, and have been removed in other browsers. Many publishers and advertisers already have strong opinions about the impact this has had on monetization of the open web.

I just can't see how this will play out well at the moment.

The scenario I've outlined is obviously extreme, but I think it's worth thinking about what the world will look like if Chrome is the only browser that supplies a technology that gives advertisers what they need.

Will the advertising industry be likely to incentivise publishers to prioritize an audience which enables them to utilise FLoC?

Maybe the landscape will become more complicated, and various competing technologies will be used by the advertising industry to maintain its ability to target adverts.

My smart TV wants me to let it target adverts based on my demographic. I choose not to. My thinking is there isn't a huge fight to convince me to do otherwise, because this threshold hasn't been broached for the majority yet.

There's a trend in modern life. We lose a liberty, and regaining that liberty is impossible in the name of profit.

It's depressing.

kuro68k commented 3 years ago

Web advertising is generally placed through a real-time bidding process, so if legit advertisers begin using FLoC extensively, then users of non-FLoC browsers will not get bid on by the higher-bidding FLoC-using campaigns and end up receiving more lower-quality or deceptive ads. Non-Chrome browsers might end up being able to give their users a better total experience by implementing the FLoC API (with additional protections, if needed) than by leaving all their users in the "null cohort".

I think it more likely that FLoC will accelerate the trend of simply blocking ads entirely. Rather than saving the ad revenue stream it will destroy it, by encouraging browser developers and users to use blocking technologies to block FLoC and any potentially low quality ads that result in said blocking.

The only solution to the problem of ad revenue going away is to create a system of safer, less annoying and most importantly privacy respecting ads. Since FLoC does none of those things it is part of the problem, not the solution.

flatsiedatsie commented 3 years ago

Exactly.

The only advertising I'm willing to accept is if it's based on the content of the page I'm visiting, not the content of my character.

We took a wrong turn somewhere around 2003, and we need to go back and develop advertising based on the content of webpages more. That way we don't need to normalise large-scale surveillance infrastructure and businessmodels. The risk for abuse is huge, while the financials gains are relatively small.

Advertising based on page content: 1% of people clicks an add. Advertising based on surveillance: 2% of people click on adds.

It's disproportionate.

After Dutch advertising service STER noticed that over 90% of people were opting out of surveillance-driven advertising, they switched to advertising based on page content. It turned out they actually made more money this way.