3 ms·
Adtech companies know exactly who is responsible. Every ad on any serious network is approved before going live. Every single auction is logged, that's just bas
by manigandham 8y ago
Adtech companies know exactly who is responsible. Every ad on any serious network is approved before going live. Every single auction is logged, that's just basic tracking and necessary for billing anyway.
The problem is that there are no serious consequences in this industry. Ad fraud isn't a technical problem, it's a business problem, one that most companies are not incentivized to solve or prevent.
- crankylinuxuser 8y agoIf companies are "people" in the legal sense, why isn't there a 'corporate death penalty' for utter gross violations that would get real humans locked in a case for decades? Is the CFAA only for fleshers? Why so?
- _jal 8y agoBecause that legal fiction is only employed when it is convenient for the owners of the fictive "people" in question.
- manigandham 8y agoI don't know about all that. The "why" is because it's a 12-figure global industry including two of the most valuable companies in the world with unlimited resources and lobbying power, combined with highly technical mechanics and a complex network of business relationships that no politicians have any real understanding of. I must add that there's also a (sometimes irrationally) strong reaction by those politicians and many others to working with anyone in adtech to even attempt to solve these problems which clearly doesn't help.
- halfjoking 8y agoI work for a publisher and basically these ad networks say "you'll lose 50% of your revenue if you implement 'safe frames' in Google AdManager 360. We have some publishers who do it but basically no advertiser wants to run ads in a safeframe." So you're saying the standard practice is to let the malicious ad buyers from your network run arbitrary javascript on our sites... "Yes that's what our publishers do..."
- manigandham 8y agoSafeFrames are fine for most ads that are isolated in a box on the page, and those that wait for user interaction before changing dimensions. The main issue today is because advertisers want viewability tracking (which safeframe has a minimal but poor API for) and they want it with their own independent providers. Also prebid.js used for header bidding does not support safe frames. Things will improve though because IntersectionObserver is now built into modern browsers and even works in cross-domain iframes. There's also the Open Measurement SDK[1] from the IAB to standardize viewability finally, and Prebid.js is also working on using SafeFrames. 1. https://iabtechlab.com/standards/open-measurement-sdk/ https://iabtechlab.com/standards/open-measurement-sdk/
- nitwit005 8y agoThis is overlooking the problem of detecting that people are doing something malicious. There is a fairly large variety of unwanted behavior, both by the advertisements, and by the sites or apps hosting the ads.
- manigandham 8y agoThat's covered under "no consequences". You don't need fancy tech, most of the malicious behavior comes from a limited set of bad actors and vectors. It's just not stopped because nobody cares to.
- nitwit005 8y agoSorry, but you genuinely do need fancy tech to detect it. There are multiple businesses that do nothing but that, plus teams at most of the big ad tech companies. You may notice this article cites Protected Media and DoubleVerify. You can't manually audit that kind of traffic volume. Several of the ad tech companies have pressed law enforcement to prosecute some of these people, but mostly unsuccessfully. They're often in places that are difficult to extradite from like Russia, hiding behind some botnet: https://adexchanger.com/online-advertising/why-the-doj-finally-cracked-down-on-ad-fraud/ https://adexchanger.com/online-advertising/why-the-doj-final...
- manigandham 8y agoMy point is that detection doesn't solve anything, it's treating the symptom. The solution is to stop working with bad actors, but nobody does that because it's easy money and there's plausible deniability by creating new companies, accounts, sites, etc. Your 2nd statement is what I'm saying: no consequences refers to the lack of oversight, regulation and enforcement in the industry. That's why there's barely any prosecution. For example, Buzzfeed did an expose last year about Newsweek/IBTimes committing ad fraud, except it was already well known by everyone in adtech. People like getting paid and they're not going to stop on their own. Btw, I personally know the founders of all those companies. If fraud was eliminated then they would go out of business. They're not interested in solving the problem even if they could.
- o_nlogn_420 8y agoThis is simply not true. I work for a top 3 major ad exchange, and there's only so much validation that can happen when you process 10s of millions of requests per second that have to leave your system in 40ms.
- manigandham 8y agoLike I said, this is a business problem. Those major ad exchanges should stop working with bad actors. 95% of ad fraud is coming from known sources and/or blatantly noticeable.