Most Facebook Ad Clicks Come From Bots, Start-Up Claims

Boy, it just hasn’t been Facebook’s summer: a high-profile dumping by General Motors was followed by a busted IPO in May, and its stock continues to sink -- trading at $21.77 at the time of writing, down 43% from the IPO price. Meanwhile, a new eye-tracking study has cast (even more) doubt on Facebook’s ability to monetize mobile traffic through advertising. But the most damaging news may come courtesy of a New York-based start-up, Limited Run, which claims that 80% of the clicks on its Facebook ads came from bots.

In a much-circulated blog post, Limited Run, which offers Web site services to artists and musicians, informed fans that it is deleting its Facebook page and moving over to Twitter, with the following explanation, which I have taken the liberty of re-posting at length:

“Unfortunately, while testing their ad system, we noticed some very strange things. Facebook was charging us for clicks, yet we could only verify about 20% of them actually showing up on our site. At first, we thought it was our analytics service. We tried signing up for a handful of other big name companies, and still, we couldn't verify more than 15-20% of clicks. So we did what any good developers would do. We built our own analytic software. Here's what we found: on about 80% of the clicks Facebook was charging us for, JavaScript wasn't on. And if the person clicking the ad doesn't have JavaScript, it's very difficult for an analytics service to verify the click. What's important here is that in all of our years of experience; only about 1-2% of people coming to us have JavaScript disabled, not 80% like these clicks coming from Facebook. So we did what any good developers would do. We built a page logger. Any time a page was loaded, we'd keep track of it. You know what we found? The 80% of clicks we were paying for were from bots.”

Now, 80% seems like an egregiously high proportion of traffic coming from bots, and it may very well be that Limited Run had spectacularly bad luck (or was being fraudulently attacked by a competitor aiming to run up ad costs) – thus making it an anomalous case. But what if it isn’t? All those commonly-heard complaints about Facebook advertising “not working” would suddenly start to make a lot more sense.

What needs to happen now is a wider investigation by online ratings firms, advertising and media agencies, and any interested (and competent) tech firms and individuals, to see if they can, in good empirical fashion, replicate Limited Run’s results. Certainly nobody can deny that Limited Run’s anecdotal evidence is food for thought.

Tags: social media
Recommend (9) Print RSS
8 comments about "Most Facebook Ad Clicks Come From Bots, Start-Up Claims".
  1. Steve Kavetsky from AgooBiz, Inc. , July 31, 2012 at 7:08 p.m.
    Wow! This is what happens when a company doesn't have a solid business plan. First off, FB was NOT created to be a business. After gaining a ridiculous number of users, FB realized it needed to become profitable. That's when FB lost its way and its business plan began demonstrating multiple personalities. They did not get their act together before their IPO which was a huge error in judgement. Their exec. [sales or marketing] who recently categorized FB as Middle-of-the-funnel [sales] was motivated by crisis mgmt. more than anything else. Here are a few of the many issues FB has not satisfactorily addressed: - the released financial figures for their IPO were in question - the large percentage of FB users who didn't create their own accounts and are not active - numerous privacy concerns and distribution of personal info to advertisers - With various FB updates, FB killed many third party developers whose businesses were solely created in support of FB's platform [no heads-up to a billion dollar industry from FB about updates - they just happened] - This new allegation that 80% of ad clicks might be done by programs or bots. This latest allegation, if found to be true, can spell disaster for FB's already tainted reputation. Why? Because FB uses their ad click stats to reel in prospective advertisers [80% would be more than an honest mistake - it would be criminal]. Did GM figure something out that none of the rest of us businesses know about? With everything above being said, I still feel FB can improve and succeed. The feedback for improving FB is out there. Millions of users have not given up on it despite its problems. FB has plenty of time, resources, experts to help them fix most of their issues. What is FB waiting for? There was a time when FB [with good reason] was drunk on its own unprecedented initial success. But how long should the honeymoon period last? Steve Kavetsky Co-Founder/Pres. AgooBiz.com // The Social Commerce Network "WE work greater than me"
  2. Jim Burnette from Free All Music.com , July 31, 2012 at 7:50 p.m.
    Clicks on Ads are just a CTR. CVR is what smart Brands are measuring. Fans, Video Views, Installs, Shares, Photo Views are the true measurement of engagement on Facebook. Brands need to define the Action and Optimize. Build a compelling creative & media plan and users will Engage.
  3. Sadie-Michaela Harris from iLocal Search , August 1, 2012 at 8:02 a.m.
    Interesting post thanks for sharing, is there any substantiation to go along with it or any other possible explanation for your experiences? As you say 80% is high bot ratio!... through the roof really! To this end I'm wondering if other people are experiencing the same issues, or have the same doubts. :)
  4. Steve Kavetsky from AgooBiz, Inc. , August 1, 2012 at 12:51 p.m.
    Hi Jim - nice to meet you. You're absolutely right about what true customer engagement is. Your advice about what smart brands should be measuring is spot on. The problem with CTR as it applies to this case is that FB is charging for CTR's. If your company is getting charged per CTR and you learn that 80% of what you're paying for is fake [i.e. bot created]- then the company that's charging you is selling you snake oil rather than real results. Hi Sadie-Michaela - pleasure to meet you as well. Unfortunately, few companies have the resources that "Limited Run" has to build their own analytic software. If they didn't test it themselves, we would not even have a clue about the 80% [we both know FB would never release these figures themselves because it would be corporate suicide]. I'm hoping it is an error because if it's not, then it means a ton of companies were & still are being bilked by FB. Let's see if "Limited Run" gets a public response or if other companies figure out the same thing and come forward. Time will tell. Steve Kavetsky Co-Founder/Pres. AgooBiz.com // The Social Commerce Network "WE work greater than me"
  5. Andre Szykier from maps capital management , August 1, 2012 at 1:24 p.m.
    If this is true, then FB has a real revenue problem. I would like to give some input about ad metrics and the experiences we had in social media using tracking scripts from the various vendors such as Google, GetClicky, Quantcast,..., and Comscorey. When we posted content to various social media sites, the referral link went to our analytics server before going to the target url of the content. As part of this process, we had our own php tracking script plus scripts from several ad metric vendors on the php landing page. We figured that we should get one to one conformance on most of the ad metric variables including originating IP, browser type, OS, and so on. Not so. In fact the discrepancies between unique visitors that we showed and unique visitors shown on the content page had wide variances, both within the same vendor script and across vendors. Google Analytics had the largest variance. On some days it was as much as 60 percent! It probably has to do with how long a person stays on a page before a new session is set by the tracking script. GetClicky seemed to have the least variance but under reported clicks against our internal tracking script. Quantcast seemed to have good and bad days. On certain days it was exact and on other days showed NO traffic. Sounds like a server database issue. As for bots, we did notice that traffic of certain social media sites came from multiple ip addresses octets). For example, 50.97.149.178 is for Tumblr.com running on SoftLayer Technologies in Dallas, TX, United States. Yet, when we examined traffic from Tumblr readers, there were multiple ip addresses, sometimes from the Tumblr user and sometimes ip octets in SoftLayer ISP. I would be interested in how the bot ip addresses were categorized by Limited Run. This might explain the 80% estimate which could be much lower in reality.
  6. Steve Kavetsky from AgooBiz, Inc. , August 1, 2012 at 2:01 p.m.
    Hi Andre. You provide a concise, detailed, technical explanation for Limited Run's 80%. If it is as you stated, then that's great news and it means there's less to worry about. What bothers a lot of users is that when Limited Run tried to contact FB's customer service about their findings, they received NO RESPONSE. Is this how a business should treat its paying customer? If they had received an explanation such as the one you provided here, then this would not even be a story worth writing about. Steve Kavetsky Co-Founder/Pres. AgooBiz.com // The Social Commerce Network "WE work greater than me"
  7. Andre Szykier from maps capital management , August 16, 2012 at 4:25 p.m.
    Response to Steve Kavetsky. I agree, that FaceBook is less than forthcoming on how sophisticated or not they are in their network engineering. They probably are trying just to keep their massive infrastructure running 24x7. A lot of our social algorithm development focuses on analyzing referral traffic from social media sites to third party publisher content. As more traffic referrers are now using shortened URLs (example Bit.ly), I suspect that this bot issue will become more visible. Perhaps the services that shorten URLs can investigate as we have done, whether Limited Run is correct in its claim for other sites.
  8. Simon Harris from Affordable Media , September 19, 2012 at 9:54 a.m.
    The issue of fraud traffic – and in particular bots – within the media buying business is staggering. Experiences at UK tech start-up TrafficCake.com - http://www.trafficcake.com - showed even reputable ad-networks delivering 30%+ fraud traffic, whilst some of the smaller traffic sellers, particularly those targeting small affiliates and internet marketers unashamedly send 100% bot traffic. They continue to operate with impunity, often from territories beyond the grasp of authorities. When their website or Pay-Pal account is shut down they simply open up new identities. Traffic buying needs to be more safe and transparent. Knowledge is important – traffic buyers need to know what steps they can take to protect themselves. And don’t forget if 30%+ of all online marketing ad spend is fake – that’s a lot of money going missing from legitimate parts of our economy.