A Facebook bot purge clobbered USA Today | Hacker News Reader

A Facebook bot purge clobbered USA Today

cjr.org

152 pointsr72171 comments

Instead of crazy conspiracy theories, how about this: USA Today isn't actually that popular, but some botnet programmer with an imperfect understanding of American culture thought it was a brand ordinary Americans all love and read regularly. So they were all instructed to follow their page in an attempt to make the bots look legit.

USA Today has the highest circulation in traditional newspapers. Not really sure why that is, but my guess would be that they have some enduring relationships with hotel/motel chains. It's definitely a brand that resonates.

Source: https://www.infoplease.com/arts-entertainment/newspapers-and...

USA Today was the first newspaper of its kind. Flashier, more concise, graphic-laden news. A larger focus on celebrity, entertainment, etc. news. From wikipedia on USA Today: "derided by critics, who referred to it as "McPaper" or "television you can wrap fish in," ". It was the Buzzfeed of its time.

I never really thought of that but that is a fantastic analogy.

USA Toady: for people who think television news is too complicated.

It's a generic newspaper without a regional focus, unlike the NY Times, LA Times, etc. It's distributed nationally, and it's popular with travelers and companies who cater to travelers (airlines, airport kiosks, hotels, etc).

Our city's paper focuses exclusively on local and regional issues and includes a USA Today section to cover national stories. I suspect a lot of smaller metros' papers do the same.

Specifically those owned by Gannett: https://en.m.wikipedia.org/wiki/List_of_Gannett_Company_asse...

Smaller/local papers also make extra money on the side through print contracts -- USA Today pays the local paper to print USA Today for distribution in that area.

Thirty years ago, I traveled a lot for work. In hotel rooms in Springfield, Missouri, or Dubuque, Iowa, I would often enough find USA Today slid under the door when I got up. Now, USA Today didn't really stack up against the local papers in Chicago, New York, or Atlanta, but a smaller city's paper could be pretty thin, and with national coverage largely pulled from the wire services.

I tend not to read it much, but it would serve for breakfast reading in a small town.

And airlines.

That was my first thought as well. Bots just looking for a page with "USA" in it.

This is what immediately crossed my mind too. I have talked to quite a few people from other countries who assumed that USA Today is our foremost national paper, like the American Der Spiegel.

I have no idea where the idea comes from though.

Because it's ubiquitous and small enough to conveniently carry. Source: I visited USA for the first time recently.

It's also of a much higher* quality than I expected. It's got something for everyone.

* my expectations were very low

I went to American University and it has a huge number of international students, both because of its location in DC and also because of its name, they thought it was the national university.

I have no idea where the idea comes from though.

It is one of the only newspaper of note without a city name in the title. Also it is one of the best selling newspapers in the US.

My thoughts too. It's also something that might come up if you were searching for "USA" pages that had lots of likes already.

Ok, why is that more plausible than USA Today buying fake page views? Why is that such a crazy conspiracy theory?

They reported the unexpected activity from foreign countries to Facebook themselves. I don't think they were in on it.

> The purge of such accounts culminated a six-month effort by Facebook to combat the scam, which routed faux profiles through Bangladesh, Indonesia, Saudi Arabia, and other countries.

> ...

> The USA Today spokeswoman told CJR that they flagged the issue for Facebook after noticing an unusually large uptick in followers from the aforementioned countries. “Since we first brought this issue to Facebook’s attention, we have been in close communication with them and look forward to a swift solution that prevents this illegitimate activity from happening on our Facebook page in the future,” Maribel Wadsworth, Gannett’s chief transformation officer, told USA Today Friday.

When sites pay other services to increase their fb likes, they are always from other countries for some reason. Because in this case the payer only wants to see the actual like count go up.

Not saying we have proof that they did this.

One stakes the reputation of an entire company, and one is as easy as a script kiddie setting up a bot.

A lot of people seem to think that setting up a botnet is difficult, and it's very surprising when you decide to research and actually do it. It's ridiculously easy. You can rent out established botnets with almost no effort, and if you can run payloads on compromised internet-enabled devices, you can set up your own and coordinate it without much fuss.

This is a big problem that no one really knows how to solve.

I wonder if there's a way to build a bot net that takes down other bot nets. Basically, create a feedback loop such that the decreasing price of a bot net leads to some action that drives the price back up.

I mean, look at the shenanigans Wells Fargo gets caught up in, which are worse than "buying likes" bot activity. Is it that unbelievable some manager with a bonus tied up in vanity metrics would use bots to boost them?

because it's less depressing.

I'm proposing a corollary to Occam's razor called avs postulate:

1) The likelihood of an explanation being true is correlated with how boring it is

2) crazy conspiracy theories are a useful exercise with which to guide convergence and formal presentation of evidence of 1.

Because that is an illegal / unethical business practice which would severely tarnish their reputation?

Assuming they even know what they are doing.

I'm perplexed here how there doesn't seem to be talk of the fact Facebook had 5.6 million bots disappear in one purge.

In almost every Twitter thread on here there is plenty of disparaging talk about bots and how it makes the social network's MAUs distorted or make it less valuable. Yet, here is quite obviously an indication that Facebook suffers from the same-- if not worse problem because it's a walled garden and less easily viewed.

If anything I'd suspect that an inordinate amount of users on Facebook and its MAUs are bots because of this... that the problem may in fact be more prolific than twitters but goes unnoticed because we can't browse Facebook like we can twitter. It would certainly be a more valuable target as a result since people falsely believe in Facebook's metrics-- versus, say Twitter, where advertisers can see the bots.

... I'm just befuddled right now that the conversation is about how the bots liked a USA Today page and not the bots themselves. Maybe I missed the previous discussion. If I was USA Today, I'd certainly consider dumping any effort for Facebook as the numbers and metrics cannot at all be trusted to be real.

When I did analytics at MySpace, we had a concept called "one hit wonders" for new users that hit one page view per month. That was the typical behavior of a bot for MySpace at the time. Internally, the "one hit wonders" filter would be used by analysts. Externally, the filter would not be used because there was a desire to inflate stats.

As a side note, the system to create these stats was pretty cool for its time. It was a 52 node asterdata cluster on bare metal (MySpace had deep pockets at the time). It could analyze every session ever created on MySpace in the span of a few hours.

I fully agree. It's different on Twitter because it is far more public. Plus things like Retweets cause the problem. With Facebook it isn't much a problem except for more public entities like USA Today. On Twitter however I don't understand why they don't take more action. I know one account that is single handedly responsible for 99.999% of the spam I see. And it shouldn't be that hard to find and detect.

Facebook reported 1.23 billion daily active users as of last December [0]. 5.6 million bots is only 0.46% of the "user" base, so in terms of raw number of users it doesn't really seem all that bad.

I would be curious, however, how much engagement that 0.46% accounted for. Didn't seem like much since the article mentions USA Today not seeing much of a dip in engagements.

[0] https://newsroom.fb.com/company-info/

This was only one purge tipped off by USA Today.

I would guess that Facebook has 10-20x that number of bots total, since this appeared to be a single network that hadn't become "active" (as per the story).

A quick Google puts Twitter at 9-15% bots, which makes the estimates roughly line up with what I'd expect: Twitter is slightly worse (2-3x bot concentration) but the two aren't incomparably different.

My wife just experienced something that makes me assume Facebook's bot/abuse algorithm is malfunctioning.

She just tried to "sell" something for free on Facebook Marketplace. Due to the popularity, she received 100+ messages in a minute -- the notifications were rolling on her phone.

Shortly thereafter, Facebook locked her out of her account for 72 hours and asked her to verify her identity. I guess they thought she was a bot? Which is strange because why would a bot want to receive 100 messages rather than sending them?

Why do you think she was flagged as a bot? I think it's more likely she was flagged as a scammer. To receive such a high volume of inquiries suggests something unusual going on. In her case it sounds as innocent as giving away an item that is in high demand, but in many cases a similar pattern might be a deal that's "too good to be true."

To phish people. Bots (like on Tinder, dating websites, selling websites, etc) will make the same post to a ton of different groups, and then send the people who message them to phishing links.

Ugh, lockout must be pretty frustrating. Receiving one message could be worth as much as sending 1000 messages when valued by conversion to the spam end-goal. It's also a less obvious technique and one that's more difficult to block. Clearly the algorithm is not as precise as it should be, but I'm sure this mechanism is preventing a lot of abuse even though it blocks some normal use.

Ok, so I got locked out of my account last night for 72 hours. I have no idea why. I had to upload an image of myself for their team to verify. I wasn't doing anything odd whatsoever. Still waiting for my account to be reactivated.

Maybe they just wanted a picture of you?

Hey, it's less awkward than asking him what he's wearing

One can dream!

You're beautiful.

In that situation I'd probably giving my best "grumpy old man from Up is not amused" face for the camera to quickly get across that 'yes, it's me, now let me back in' or such.

You could use a Facebook bot as CC for a botnet. Such bot would initially receive a lot of messages without necessary send out their own, and also being something Facebook would want to ban.

A bit late to this party, but ALL machine learning algorithms are wrong some fraction of the time. It's not malfunctioning, it's a false positive.

What was the product?

Maybe I'm jumping to conclusions, but it seems to me like they had engaged some "social media marketing" firm that was using these like bots as part of their tool kit and were probably getting paid by USAToday on a performance basis.

Actually I think it hurts them to have a large number of inactive followers.

> While USA Today’s horde of fake followers did not appear to “like” or comment on posts at a significant clip, the presence of so many of them could serve to muddle the rate at which the audience was engaging with content.

Yes, a poo follower to share/like ratio actually damages the reach of your posts.

That made me wonder if it could possibly be a competitor inadvertently taking control over their page by saturating it with inactive followers.

Then again if the accounts themselves are inactive in general, not just on this page, maybe it's something Facebook accounts for… maybe not… hard to say.

If I recall correctly they did use Demand Media content so they were at least, assuming I'm remembering right, somewhat interested in shady seo tactics. Disclosure: used to work for Demand media.

for a fantastic expose of how facebook is riddled with fraud and bots, check out veritasium's two videos:

facebook fraud https://www.youtube.com/watch?v=oVfHeWTKjag

the problem with facebook https://www.youtube.com/watch?v=l9ZqXlHl65g

I never like anything on FB; I've always assumed it a pointless waste telling no one anything useful since it is so (apparently) easy to scam. I wonder if you should have a certain level of activity for a reasonable period of time before you are counted for anything real.

I wonder if this was somehow an attempt to censor USA Today? This many followers that don't engage decreases the odds of their posts showing up in users' news feeds.

Fun conspiracy though it's probably more likely that they were a seed in a botnet algorithm.

Okay, on the face, this looks bad... What are some legit reasons this would happen? Other than a marketing person going insane over metrics, why would you bother?

USA Today could be a credible source for bots to follow for free to increase their own legitimacy.

How does it increase the legitimacy? Page likes aren't public (when viewing from/as the page).

So this is about Twitter and not FB, but I went on a weird trip down the rabbit hole when looking into how certain accounts were becoming popular/growing so quickly.

There are a whole bunch of "networks" of fake accounts, often complete with very legitimate looking photos, names, and even posts. The main difference, however, is they follow and retweet a lot.

If you look at what they follow and retweet, however, well over half seems to be "legit" stuff that wouldn't be getting involved in such schemes. I think the reason is that if Twitter decides to crack down on bot networks, they might look at what the bots are following and promoting to work out who to penalize.. except they aren't going to start deleting CNN, Fox News or random celebrities.. so they really have no great way to tell who to punish. I suspect FB bots work in a similar way.

Not the legitimacy of the page being liked, of the bot's profile. Filling it with dummy information, including various liked pages, posts, etc, to use them for astroturfing.

People follow pages and join groups and all that, so bots have to do that as well to blend in. It's to make the bot more legitimate.

I'm not sure they're bots, rather people cheaply paid to create fake accounts, or scammers. I've seen many of them, and some of them are quite amateurish. They would say they went to Florida University and come from Florida City, working in... oil.

Maybe they paid some blackhat "marketing firm" not fully realizing it?

What makes you think they didn't realize it?

"The USA Today spokeswoman told CJR that they flagged the issue for Facebook after noticing an unusually large uptick in followers from the aforementioned countries. “Since we first brought this issue to Facebook’s attention, we have been in close communication with them and look forward to a swift solution that prevents this illegitimate activity from happening on our Facebook page in the future,” Maribel Wadsworth, Gannett’s chief transformation officer, told USA Today Friday."

The article suggests USA Today brought this to Facebook's attention. If USA Today had been doing this on purpose, even indirectly, that would be a bizarre move. We also have nothing from Facebook contradicting this claim.

You can hypothesize nobody asked Facebook about it or that USA Today is now lying to try to save face, but those would be hypotheses that are possible, but not currently supported by the facts. It is a reasonable interpretation that USA Today is indeed a target of some sort and not the instigator. I am also not sure exactly what the bots sought to gain from this, but I've seen enough similarly crazy things that made sense once an explanation came out that I'm willing to give some time for such an explanation to come out.

I've got no love lost for the media but they're still shining beacons of virtue compared to the people authoring and running bot networks for this sort of thing, so I don't find it that hard to trust USA Today enough to consider their version to be the most likely story. (Not the only story, but pending further data, the most likely one.)

Call me cynical, but I'm betting USA Today paid for millions of fake followers to pump up their stats (making them a more attractive platform for native advertising), got wind that Facebook was onto them and decided to break the story first and act like they were the ones to tell Facebook about it in the first place to save face. Nowhere else besides this woman's quote does it say that Facebook initiated their bot-destroying campaign after USA Today told them about all of their fake followers. They're lying, plain and simple.

And this is based on... what, exactly?

1. They didn't break the story, CJR did.

2. Why would they buy so many likes? In my experience, native ad sales revolve around native metrics. Buying bots to like your page (but not visit or interact with your website) won't help you there.

It's based on being cynical due to working in marketing for 15 years. Obviously I'm only speculating. I mentioned native ad sales because sites that engage in native marketing typically sell a package that includes posting the native content to their Facebook page, which has X number of followers. The higher that number, obviously the more attractive it is.

Yeah, I don't get people's weird obsession with trying to explain away blackhat marketing when it's perfectly normal and acceptable to put your own interests ahead of Facebook (assuming you aren't breaking a real law).

Because they may have hired an Agency that hired an expert Agency that hired a Freelancer who hired an remote personal assistant to do the job.

Here's a crazy though. A competitor hires someone to create a botnet to destroy them. You have the fake accounts start liking shitty content and ignoring good content. If the targeted company followed the desires of these fakes they would eventually go out of business.

It sounds like of no fault of their own, USA Today was a god FB resource to like and shape in order to build fake profiles.

Now I understand how Facebook as 1.4 Billion monthly active users.

Headline sounds interesting, any sources without interstitial ads? I saw one on this article and obviously left immediately never to return.

Reply on news.ycombinator.com