First Victory of the Mobile Ad-Blocking War: A “L.E.A.N.” Digital-Ad Standard | Hacker News Reader

First Victory of the Mobile Ad-Blocking War: A “L.E.A.N.” Digital-Ad Standard

theatlantic.com

23 pointsForHackernews15 comments

I was interested in this standard until I read the following:

> These ads will still track users across the web, though. Cunningham notes that the IAB will alter its general industry standards so that ads appear “before, but never AFTER” someone makes a purchase (emphasis his).

Just don't track users. Don't mine data about your readers. It's impersonal, it's ugly, and it's wrong. What I want on the web is for you to create the content you want to share, and I want you to pick the advertisements you think I'll like. I don't want you to delegate that responsibility to a tracking third-party.

A few years ago I pulled all the analytics services out of my apps. All unnecessary server contact was removed (i.e., anything not initiated by the user). I was just emailed by a group wanting the "backend usage data" from one of my apps which has been very popular with kids; the group specifically wanted it for that reason. I was very proud to be able to tell them that the app makes no contact with any servers whatsoever.

>Just don't track users. Don't mine data about your readers.

I'd expand this to: "Don't ask your users to trust you or anyone you have a business relationship with. Be absolutely transparent about what they're giving up for the content you're providing." --Edited for clarity

I went to the Sydney Morning Herald on my OS X box the other day, and found that after about 30 seconds I could no longer scroll up and down the article. Turns out they had some big interstitial ad that was blocking the down and up buttons. I had a look at the JavaScript running in the background and counted at least five trackers all running at once.

This has honestly now changed my behaviour. I've decided that the SMH is too slow for too little value for the already paucity of quality news and analysis it provides. Heck, I can't even post comments to their site any more - it all gets rejected after waiting in a moderation queue for ages. I cancelled my subscription and now I read my news from The Guardian Australia which has more content and better analysis, and certainly very few ads!

Goodbye SMH, you were an institution for so long, but your golden days are behind you and you are becoming increasingly irrelevant.

This experience has always been why I've been against ad blockers: in some sense it 'rewards' bad publishers who choke their site with ads.

I've often wondered whether a similar policy should be given where you get the ability to whitelist certain third party domains and prevent any AJAX requests from triggering (and blocking) to domains outside of the website you are browsing.

The only reason ad blockers work is because they serve external resources. When media sites really wise up, they will figure out a way to serve ads from their own domains. Building a system for ad bidding, rotation, tracking, advertiser monitoring, and external auditing is too big of a task for many sites, so I think there's a startup opportunity to develop and sell a platform that can easily be installed on a server.

Can you give me extra details on this topic. I mean some extra information and resources.

There is not much to say. Back in the "old days", you had article with text and images (related to the article, e.g. images/1.jpg, images/2.jpg) and somewhere in the middle of article was banners/123.jpg, but it could easily be images/3.jpg. This was mostly good because both parties (content provider and content consumer) both pay the same price for the bandwidth. This keep ads on the sane levels because too much add would be costly for provider so they keep them moderate.

Then something changed and ads was outsourced to third parties, so instead of images/3.jpg you would have link to external domain, e.g. http://ads.com/3.jpg. Because this way content provider no longer pay for ad bandwidth (it is hosted on other server) they could literally add infinite amount of ads. 8MB jpeg? no problem 3MB javascript? no problem, 4MB animated gif? no problem. This will chew through your data plan very fast so users started blocking ads, the simplest way is block content delivery networks of all ad companies.

Next logical step (for displaying ads) is to move them back to the original domain of content provider. Then you cannot tell which of then images/1.jpg, images/2.jpg or images/3.jpg is ad, unless some human will mark it. This is difficult, prone to errors and abuse and would make ad blockers gigantic.

Bandwidth is not the only reason to block ads. Removing distractions is more important for me.

It's possible (or should be possible, it was/is in the browser iCab) to block images depending on their size (and ads often are of the same size, e.g. banner ads). Sure, the browser has to start loading them, but after reading an image's metadata it can stop. And if the img tag contains the image's dimensions, even that can be skipped.

They already have... My employer are running multiple native advertising techs to alleviate the problem of ad blocking. We are also running anti-ad-blocker techs to measure and take action on users with active ad-blocking technologies.

Personally, I'm thoroughly against this arms race where both we and our users are loosing. But the financial side of our company sees it a different way.

You're talking about 3rd-party tech? Could you name one (or some?). Given that Linux is big server-side the tech would have to fit into that stack so I'd like to see how that's done for curiosity sake.

I think having it native is actually better because then companies will see the impact on their bandwidth and compute resources so that'll force ad-tech to be less bloaty. In theory.

Ad's are blocked at such sites by URL path (for images) or by DOM path, or by regex. It is even easier and predates adblock, see Greasemonkey plugin.

You're thinking of domain based adblockers. There's also DOM modifying adblockers and javascript manipulating adblockers.

Why would spam detections techniques that we use for email not work for content?

Reply on news.ycombinator.com