Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Company withdrawing from Facebook as analytics show 80% of ad clicks from bots

facebook.com

Post reply on HN
Print view View original

211–220 of 381 posts

Re: Company withdrawing from Facebook as analytics show 80% of ad clicks from bots

#211

Post by stef25 »

samstave wrote: post #112
reustle wrote: post #49

Copying the text just incase FB decides they don't like it --- Hey everyone, we're going to be deleting our Facebook page in the next couple of weeks, but we wanted to explain why before we do. A couple months ago, when we were preparing to launch the new Limited Run, we started to experiment with Facebook ads. Unfortunately, while testing their ad system, we noticed some very strange things. Facebook was charging us…

That escalated quickly! And so it seems that Facebook could be proven to be the new groupon. I would absolutely not be suprised if the hits were FB bots - however I'd expect they wouldn't be that directly stupid so as to have an Ad-bot net in house. Surely this is some service they outsource for plausible deniability. Something they would have learned from their relationship handing all user data over to the NSA/CIA.…

What user data stored on FB servers could be useful to intelligence agencies of that level? People brag about crimes they committed, IP addresses are logged and people may "like" jihad but criminals targeted by the NSA are surely not at that level of stupidity.

Also, FB may not have much of a choice to comply with those agencies, or find themselves pissing off a lot of high level people by refusing to comply with government agencies. But outsourcing / tolerating a network of bots to vastly boost revenue is probably considered fraud, no?

Re: Company withdrawing from Facebook as analytics show 80% of ad clicks from bots

#212

Post by zzzeek »

th0ma5 wrote: post #204
zzzeek wrote: post #195

Earlier quoted context omitted.

this is a great use case for bayesean filtering, and I bet you could even hook up an existing bayes engine to this problem space, if not just write one up using common recipes - I've done great things with the one described in "Programming Collective Intelligence" ( http://shop.oreilly.com/product/9780596529321.do ). Things like "googlebot", "certain IP range", "was observed clicking 100 links in I'm sure facebook en…

Well yeah, getting a training data set, and then having your algo not be poisoned, seems like it is a challenge all around!

The algo can't be poisoned, if I understand the term correctly, if you only train it manually against datasets you trust.

Each iteration of training that produces improved behavior can then be used to generate a new, more comprehensive training set. When I wrote a bayes engine to classify content, my first training set was based on a corpus that I produced entirely heuristically (not using bayes). I manually tuned it (about 500 pieces of content) and from that produced a new training set of about 5000 content items.

Eyeballing it for a couple of months initially, manually retraining periodically, spot checking daily as well as watching overall stats daily for any anomalous changes in observed bot/non bot ratios will work wonders.

Re: Company withdrawing from Facebook as analytics show 80% of ad clicks from bots

#213

Post by slig »

logn wrote: post #202
nhashem wrote: post #162

My startup is essentially an advertising aggregator (pooling traffic from a variety of publishers and routing it to advertisers) and dealing with things like bot detection is a HUGE chunk of what we work on, technology-wise. Let me try and give you an idea of how deep the rabbit hole can go. - Okay, you want to detect bots. Well, "good" bots usually have a user agent string like, "Mozilla/5.0 (compatible; Googlebot/2…

could you just add a mouse-over event to each ad that sets an isHuman flag?

IIRC, years ago, Google changed one parameter of it's AdSense click url based on the number of times the user hovered the ad.

Re: Company withdrawing from Facebook as analytics show 80% of ad clicks from bots

#214

Post by SoftwarePatent »

seiji wrote: post #98
hobonumber1 wrote: post #94

Earlier quoted context omitted.

This is besides the point.

Is it though? It seems they spent a lot of time throwing more and more javascript at the problem until they realized they couldn't figure out what was happening. If they had real logging from the beginning, it would have unconfused them much sooner.

thanks for coining the word "unconfused"

Re: Company withdrawing from Facebook as analytics show 80% of ad clicks from bots

#215

Post by ZachPruckowski »

logn wrote: post #202
nhashem wrote: post #162

My startup is essentially an advertising aggregator (pooling traffic from a variety of publishers and routing it to advertisers) and dealing with things like bot detection is a HUGE chunk of what we work on, technology-wise. Let me try and give you an idea of how deep the rabbit hole can go. - Okay, you want to detect bots. Well, "good" bots usually have a user agent string like, "Mozilla/5.0 (compatible; Googlebot/2…

could you just add a mouse-over event to each ad that sets an isHuman flag?

That would break screenreaders, some touchscreen mobile devices, and probably other things I'm not thinking of.

Re: Company withdrawing from Facebook as analytics show 80% of ad clicks from bots

#216

Post by nerdo »

logn wrote: post #202
nhashem wrote: post #162

My startup is essentially an advertising aggregator (pooling traffic from a variety of publishers and routing it to advertisers) and dealing with things like bot detection is a HUGE chunk of what we work on, technology-wise. Let me try and give you an idea of how deep the rabbit hole can go. - Okay, you want to detect bots. Well, "good" bots usually have a user agent string like, "Mozilla/5.0 (compatible; Googlebot/2…

could you just add a mouse-over event to each ad that sets an isHuman flag?

The bot could just fire the mouseover event.

Re: Company withdrawing from Facebook as analytics show 80% of ad clicks from bots

#217

Post by pbhjpbhj »

reustle wrote: post #49

Copying the text just incase FB decides they don't like it --- Hey everyone, we're going to be deleting our Facebook page in the next couple of weeks, but we wanted to explain why before we do. A couple months ago, when we were preparing to launch the new Limited Run, we started to experiment with Facebook ads. Unfortunately, while testing their ad system, we noticed some very strange things. Facebook was charging us…

>in all of our years of experience, only about 1-2% of people coming to us have JavaScript disabled //

I use noscript. When I enable it for visits it's usually by enabling the domain itself and leaving third-party scripts disabled. Depending on your area this could be significant.

Also the possibility of blocking of the scripts by other elements not loading - for example if the browser can't parallelise the requests for some reason and a preceding request can't be handled then the page might be displaying well before the script is loaded, especially if it's loaded at the bottom of the markup.

Re: Company withdrawing from Facebook as analytics show 80% of ad clicks from bots

#218

Post by EricDeb »

A related anecdote: My FB account was compromised recently (using my trivial 5 character password) and all the attackers did was occasionally sign on and like random pages. I kept getting odd company posts in my newsfeed and eventually traced the problem. These phishers/advertising bots are getting more and more creative...

Re: Company withdrawing from Facebook as analytics show 80% of ad clicks from bots

#219

Post by jyu »

This whole thing seems fishy to me. Facebook is one of my best sources of traffic. It's cheaper than AdWords, I can do keyword, demo, and geo targeting all built in. And if I have a hit campaign, it's pretty easily scalable. If you tie it into specific performance, like $ / email sign up, or $ / pageview, it blows away adwords. Getting in touch with someone from Facebook can be a pain, but it's drastically easier than AdWords or other ad networks.

This whole thing seems like one big linkbait/PR play. It's a really big claim from a company that doesn't even have an about us page, has only 500 Facebook likes, and basically has 0 traffic.[1]

I'd like to know if they have tried other paid traffic sources, or have concrete evidence. Inexperienced advertisers tend to spend $50-$100, expecting instant results. And when it inevitably doesn't work out, they blame anyone and everyone.

I don't agree with the second point. Facebook isn't "holding your name hostage" and being "scumbags". You didn't sign up with the right name from the beginning, and now you're complaining. Instead of taking 5 minutes to sign up for a Facebook Page with name@limitedrun.com and get the page you want, you decided to sling mud with no real evidence. With only 500 facebook likes, the switching cost is low.

[1] http://siteanalytics.compete.com/limitedrun.com/ https://www.facebook.com/limitedpressing http://who.is/whois/limitedrun.com/

Post reply on HN