Slashdot Mirror
How Much Internet Traffic Is Fake? Turns Out, a Lot of It, Actually. (nymag.com)
Long-time Slashdot reader AmiMoJo shared this article from New York magazine:
In late November, the Justice Department unsealed indictments against eight people accused of fleecing advertisers of $36 million in two of the largest digital ad-fraud operations ever uncovered... Hucksters infected 1.7 million computers with malware that remotely directed traffic to "spoofed" websites.... [B]ots "faked clicks, mouse movements, and social network login information to masquerade as engaged human consumers." Some were sent to browse the internet to gather tracking cookies from other websites, just as a human visitor would have done through regular behavior. Fake people with fake cookies and fake social-media accounts, fake-moving their fake cursors, fake-clicking on fake websites -- the fraudsters had essentially created a simulacrum of the internet, where the only real things were the ads.
How much of the internet is fake? Studies generally suggest that, year after year, less than 60 percent of web traffic is human; some years, according to some researchers, a healthy majority of it is bot. For a period of time in 2013, the Times reported this year, a full half of YouTube traffic was "bots masquerading as people," a portion so high that employees feared an inflection point after which YouTube's systems for detecting fraudulent traffic would begin to regard bot traffic as real and human traffic as fake. They called this hypothetical event "the Inversion...."
[N]ot even Facebook, the world's greatest data-gathering organization, seems able to produce genuine figures. In October, small advertisers filed suit against the social-media giant, accusing it of covering up, for a year, its significant overstatements of the time users spent watching videos on the platform (by 60 to 80âpercent, Facebook says; by 150 to 900 percent, the plaintiffs say). According to an exhaustive list at MarketingLand, over the past two years Facebook has admitted to misreporting the reach of posts on Facebook Pages (in two different ways), the rate at which viewers complete ad videos, the average time spent reading its "Instant Articles," the amount of referral traffic from Facebook to external websites, the number of views that videos received via Facebook's mobile site, and the number of video views in Instant Articles.
On Twitter the author also shared a Twitter thread by the Washington Post's director of advertising technology, who shares his own complaints about the ecosystem of online advertising. "The problem isn't just that the internet is full of fakery and bullshit and bad numbers and malfunctioning metrics and bullshitters and fraudsters. The problem is that all the fake shit is layered on top of other fake shit and it just COMPOUNDS itself... Like you get fake users, who get autoplay videos which no one is really watching....
"That's not even counting the entire ad campaigns that are fake where the product is just a bullshit excuse to collect data on you."
How much of the internet is fake? Studies generally suggest that, year after year, less than 60 percent of web traffic is human; some years, according to some researchers, a healthy majority of it is bot. For a period of time in 2013, the Times reported this year, a full half of YouTube traffic was "bots masquerading as people," a portion so high that employees feared an inflection point after which YouTube's systems for detecting fraudulent traffic would begin to regard bot traffic as real and human traffic as fake. They called this hypothetical event "the Inversion...."
[N]ot even Facebook, the world's greatest data-gathering organization, seems able to produce genuine figures. In October, small advertisers filed suit against the social-media giant, accusing it of covering up, for a year, its significant overstatements of the time users spent watching videos on the platform (by 60 to 80âpercent, Facebook says; by 150 to 900 percent, the plaintiffs say). According to an exhaustive list at MarketingLand, over the past two years Facebook has admitted to misreporting the reach of posts on Facebook Pages (in two different ways), the rate at which viewers complete ad videos, the average time spent reading its "Instant Articles," the amount of referral traffic from Facebook to external websites, the number of views that videos received via Facebook's mobile site, and the number of video views in Instant Articles.
On Twitter the author also shared a Twitter thread by the Washington Post's director of advertising technology, who shares his own complaints about the ecosystem of online advertising. "The problem isn't just that the internet is full of fakery and bullshit and bad numbers and malfunctioning metrics and bullshitters and fraudsters. The problem is that all the fake shit is layered on top of other fake shit and it just COMPOUNDS itself... Like you get fake users, who get autoplay videos which no one is really watching....
"That's not even counting the entire ad campaigns that are fake where the product is just a bullshit excuse to collect data on you."
Pond scum feeding on pond scum. I'm having a hard time drumming up concern.
Who can tell the difference?
https://twitter.com/TitaniaMcG...
Titania McGrath can.
We're getting closer and closer to that XKCD dream life: https://xkcd.com/810/
I'm not sure why the part about Facebook inflating ad display numbers is included here. That was not because of bot activity. The majority of FB traffic is consumed through their mobile apps (95% of it), and you can be sure that is not bot type activity. Facebook has gone to great lengths to prevent scraping of their website, and it is extremely unlikely that the scraping of the site would involve scrolling through a newsfeed so that an ad became visible, began autoplaying and was streamed to a bot.
Facebook misrepresented the amount of time a user sat watching an ad before they scrolled on past it, plain and simple. More than likely they were counting things like a small portion of the video still being visible on the screen as being "watched". For the difference, even by Facebook's own admission, to be off by 60-80% shows this was a misrepresentation of what it meant for a user to watch a video ad on a very large scale (including in the app on mobile platforms).
Better known as 318230.
It really is quite difficult for me to feel all that sorry about advertisers and ad sellers being upset that their precious data is wrong/overstated/contaminated. The "ad wars" [on users' eyes, ears, cpu, screen space, bandwidth, patience] are so insane now, the anarchist in me is almost happy about it.
Ooops, another site that wants to shame/annoy/warn/block me because of my ad filters protecting my privacy/sanity/bandwidth/battery/security. Hmmm....
Ads need the data on users.
The brand that takes out the ads pays for it all.
With more advanced browsers getting the ad to display, work, track is getting more complex.
What can the ad brands do?
Make the browsers and OS more ad friendly.
Make users have to view ads in some nations?
Make the browser show an ad?
Domestic spying is now "Benign Information Gathering"
Why can't the operators of these servers join a multi-publisher subscription network? Two decades ago, such a network called Adult Check was popular, founded on the principle that adults can pay for nice things. One $10/mo payment bought access to all sites that took Adult Check, and the network paid publishers per page view. This helped to alleviate the sticker shock from each website charging a separate subscription.
More recently, Google Contributor could have been that network. The biggest problem with Contributor is lack of privacy, as it shares a parent company with AdSense and DoubleClick. This means Google can use page history gathered through Contributor to infer interests of a Contributor user for use on sites using Google adtech.
Advertising is black magic and always has been. The internet has shown ad peddlers that they don't actually know all that much about advertising, but now they can measure every aspect of their ignorance, hubris, and ineffectiveness... and it freaks them out.
The moderation mechanism described in xkcd #810 already resembles that in use on various forums and Q&A sites, such as Slashdot and Stack Overflow.
1. Each newly registered user sees a page of what Stack Overflow calls "review audits". This resembles Slashdot metamoderation: does what the new user sees as constructive align with what established users see as constructive?
2. Anyone who gets most of the review audits correct has posts placed in "awaiting moderation" state. Only established users can see such a post until at least one established user upvotes the post.
3. Once a user is firmly in positive reputation/karma, the user's posts skip the "awaiting moderation" state.
Yet this hasn't led to any artificial intelligence breakthroughs on the part of the spam industry. Instead, I've noticed that spammers on forums.nesdev.com appear to be humans in low-exchange-rate countries. They search for an old post, reword it, start a discussion, and days later edit the post to include off-topic commercial links. A user who isn't paying close attention is unlikely to see this karma whoring for what it is.
The majority of FB traffic is consumed through their mobile apps (95% of it), and you can be sure that is not bot type activity.
I would argue that the other way - because so MCUH of Facebook traffic is from mobile apps, that is where most of the bot activity is likely to be from.
All it takes is an Android user logged into Facebook and some background app can have plenty of likes and other things they don't even see happening...
"There is more worth loving than we have strength to love." - Brian Jay Stanley
I've always maintained that the way to beat the panopticon companies isn't with ad blockers and privacy legislation. It's to dilute the value of the data they collect by inserting so much fake data that they can no longer sufficiently distinguish real people from the bots.
There's an apocryphal story that after the end of the Cold War, a bunch of the CIA and KGB got together for drinks. The CIA spooks lamented that theirs had been the harder job. The Soviet Union was such a closed society and had so many restrictions on travel that it was virtually impossible for the CIA to get a spy in there, whereas all the KGB had to do was drive to a town next to a military base and mingle with staff from the base eating lunch there. The KGB spooks disagreed and claimed that theirs had been the harder job. The U.S. produced so much information that it was virtually impossible for them to separate out fact from fiction. If the National Enquirer ran a story about the military working on a, or some conspiracy theorist reported the military was controlling their brain waves with weather balloons, they had to devote resources to figure out if the stories were real or fake.
Patreon shares the same incremental sticker shock issue as individual website subscriptions. Just as your New York Times subscription doesn't let you view a Wall Street Journal article that your friend cited to you, viewing a single patron-only article from each of five different publishers on Patreon incurs a charge for an entire month's subscription to each of those five publishers. The a la carte price structure of Contributor and the flat monthly fee of Adult Check avoid(ed) the problem of it being more expensive to read from multiple publishers.
The best way to identify a real human these days is to look at traffic where ads are blocked.
If builders built buildings the way programmers wrote programs, then the first woodpecker would destroy civilization.
Not if you have to pay dearly for those shitty metrics in the form of less eyeballs (potential customers) or higher advertiser's cut from sales. There is no proof personalized ads (based on those metrics) even work, let alone that they are more efficient than the "spray wide" advertisements that TV and newspapers offer.
The ISP's have many reasons to _host_, if not actively foster false traffic. They are paid on the basis of traffic, and more traffic enhances _their_ market value. It's much like the USPS approach to junk mail, and ISP approaches to spam. The recipients _loathe_ the false traffic. The businesses carrying it try to strike a balance between maximum profit of selling bandwidth, advertising the size of their service, payment for websites, overwhelmed customers abandoning them for services with better filters, and the bottom line costs for servers, bandwidth, and security.
In the case of simplistic filtering, such as IP based filters, a few legitimate mails or web access requests from clients behind a banned NAT address can anger customers and cost business contracts. I've had to cope with this from every side myself. Thorough and aggressive filters risk cutting off legitimate traffic, while casual filters that guarantee all delivery can easily flood the recipient. The lessons apply to web traffic as well as they've applied to email.
I worked in ads at Google. _Every_ ad served by Google has metrics on it. It might not be clicks (brand advertisers often don't care about that, they just want to saturate the page), but there are metrics. I.e. whether it was in the viewport of the browser, etc. TV and newspaper ads don't have any metrics at all.