Hacker News new | ask | show | jobs
by darkkindness 2959 days ago
> Publishers will target Salmela’s software if it becomes anywhere near as popular as AdBlock Plus, says Nicole Perrin, an analyst at researcher EMarketer.

I'd caution about that claim. Google pulled AdNauseam[0], a uBlock Origin extension, from the Chrome Web Store since it was fundamentally disrupting Google's business model via click fraud. (It automates clicking ads in order to create noise in user tracking.) This was far, far before it became as popular as AdBlock Plus.

PiHole takes an even more aggressive stance against ads, blackholing entire networks. It poses as much as a threat to the ad industry as AdNauseam. So I'd wager that PiHole will get shut down long before it reaches the popularity of AdBlock Plus.

[0]: https://adnauseam.io/

2 comments

I disagree with you :

1) Unlike AdNauseam, Google can't do much to get in the way of Pi-Hole. This isn't within its ecosystem.

2) Actually, while it's not great for publishers, it's not that disruptive (at that scale) because it doesn't impact the advertiser's ad spend. No request > No charge.

Now if you wanted to create something that would be quite disruptive, you'd need to combine the 2, with a twist :

You could have a headless browser in a VM (for safety) that makes some of those calls and mimic a click + follow the redirect + stay on the landing page and browse another N pages.

Not every single ad obviously because that'd be easy to detect but at random, in an erratic way. That'd be like simulating an actual user.

Throw in there some cookie dropping / reset to mess with tracking and it could have some pretty interesting effects.

The reason why it'd more dangerous? Because it'd throw off performance of those ads and ML optimizations done by various vendors. Usually you expect a click-through rate (CTR) of 0.X % to maybe 2-3% and similarly an actual conversion rate in the single digits.

On a large scale, you'd see a a CTR that either go up or stay on par, but a conversion rate that is free falling and that would have some serious consequences for both the advertiser's marketing team and the publisher / ad network.

Take it one step further and have a TOR like network of those pi-holes taking care of that both to prevent device finger-printing and maybe coordinating a specific publisher / ad network / advertisers and you could see that advertiser looking at reducing its ad spend quite drastically when performance goes down the drain. And an ad network / publisher having to do some explaining.

Some kind of "activist" bot network to force specific advertisers to reduce ad spending or simply occur costs.

1) This is true. Though, AdNauseam was originally a Firefox extension and is still available on Firefox. You can still install it on Chrome from source. Maybe I was too strong saying it was 'shut down' -- I was more referring to the fact that action has already been taken against AdNauseam when it was still small. I'm curious for when Pi-Hole gets the same kind of treatment.

2) If the goal (of PiHole and other ad/tracking blockers) is to encourage a subscription model over an advertising model for publishers, I'd argue that making ads ineffective for publishers is much more important than directly attacking advertisers' profits. Without publishers willing to host ads, advertisers will lose anyways.

> Unlike AdNauseam, Google can't do much to get in the way of Pi-Hole. This isn't within its ecosystem

Is there any reason they couldn't start ignoring system DNS resolvers for "key properties" in favour of using DNS over HTTPS to themselves "for your comfort and safety"? And use pinned keys for those DoH resolvers to stop you MitMing it.

You could maybe still transparently proxy things or firewall specific hosts, but that's a lot less straightforward and higher risk of collateral blockage.

Shut down how? It's not being distributed by some fischer-price curated app store. Even if Alphabet companies essentially hire the judicial system to shut down or arrest troublesome developers for arbitrary reasons, it wouldn't stop piHole from being distributed p2p or stop others from picking up its updates and development, right?
I was too extreme with the term 'get shut down', which is obviously impossible as you say. I meant 'get action taken against them' which I argue will happen far before PiHole picks up in popular use.
what action can be taken against them?