Will's Blog This is why you have internet filters

16Mar/104

How to Take a Bad Article to Digg’s Front Page (Case Study)

Every so called social media guru has their own secret sauce when it comes to getting on Digg. Some will tell you top 10 lists work best, others say amazing images or interesting infographics are the way to go. But no social media consultant would ever tell you to submit something bad ... unless of course you had some help from the right people.

170 Diggs? WTF?

I logged in to Digg earlier tonight and noticed the usual array of items ... odd pics, cringeworthy videos, Republican drama and typical fails. I was pleased but then I noticed the following story had 170 diggs (now 217):

Employees are Brands

I clicked the link thinking it would be something interesting but instead it was a rewrite of a Google news release on an Adsense site by someone in Bangalore, India. Normally Digg users quickly bury this crap because they realize its scam or the person's vote purchases are all brand new accounts that leave comments like "great post, please write more."

Instead this article was getting natural looking votes and people were leaving legitimate comments. Furthermore the writer claimed Google invented "Search" when in actuality Google launched four years after sites like Lycos and Infoseek. Normally someone making an error in an article, even a small error gets crucified on Digg but somehow none of the comments made light of this obvious problem.

This struck me as highly unusual considering the fact that if you aren't a power user, your quality content (not junk) will at best get 4 to 10 Diggs before it dies. So what is the secret? How does someone push this crap through the internet's smartest community?

Humble Beginnings

Our story begins with "Sneh" / @onlinesneh, a content writer in Bangalore, India. If you take a quick look at Sneh's twitter feed, you'll notice the usual output that comes from India's best writers (not being facetious but its a lucrative market over there). Data recovery, mortgage advice, top 10 lists and tech news rewrites all published and pointing links to advertisers ... you know typical outsourced content writing.

Now I have nothing against this "Sneh" person, in fact if the bio on their web site is correct, they are currently working on a college degree (which is more than I can say for a lot of people).

Sneh proceeds to publish an article titled "Employees are Brands" which basically is a rewrite of a Google corporate release talking about employee appreciation. In fact the article should actually be titled "Google Employees are Brands."

She then tweets a link to the article three times but doesn't link directly to the Digg submission, instead the links in her tweets go to the site directly.

Then What Happens?

Before this hits Digg, a network of fake Twitter accounts are already promoting the content (img1). These aren't linking to social bookmarking submissions yet but they don't need to because a second network of Twitter accounts has begun tweeting it.

The story hits Digg when it is submitted by user nikki4digg. "Nikki" joined Digg on October 14, 2009 and has managed to Digg 18000+ submissions at the astonishing rate of 120/day. Consider this for a minute, if you only Digg pictures, on the site (the fastest way this could happen since news articles require reading and videos have to load and be played) and it takes you 15 seconds to click, load, view and digg the image ... that's 30 straight minutes of repeating the same action and that's assuming there are 120 worthwhile images to Digg. When you start to think about digging articles or even short videos, (assuming you read/view them) 120 Diggs becomes a full time job.

Sorry for getting sidetracked ... shortly after submission, the digg submission is tweeted by Jasicasmo ... one of the thousands of "ghost repeater" accounts on Twitter. I'm going to use that term a lot so let me define it for you. Have you ever heard of Bob FM? Bob FM delivers music without human interaction. It simply plays friendly demographic appealing radio and you can listen or you can change the channel.

Ghost repeater Twitter accounts simply broadcast links all day long with no interaction with the audience. In addition ghost repeaters tend to follow anyone that follows them and generally have high follower numbers that equal their following numbers because other bot accounts add them because they tend to post on competitive SEO or tech keywords (ex: mortgages, apple, mlm, linux, work from home, torrents, dui lawyers).

Before I go any further, I want anyone reading this to understand that if you think its crazy to suggest a coordinated large scale effort to get Diggs ... for $600 you can get your article on the front page and it that fails they'll make a second attempt.

After the initial article is published, just one other account (that of one of the early diggers) Tweets the same link but momentum is slowly growing because a number of well connected diggers have 'dugg' it. In fact 7 of the first 10 diggs (not including the author, the submitter and an employee of InRev) have come from accounts with extraordinarily high digg counts and 100+ friends.

The article then sits for the next 10 hours before a sudden turn of events puts it on page one when a new network of ghost repeaters suddenly begin Tweeting the link. The resurgence starts when digg_worldnews (which doesn't appear affiliated with the service) tweets with a link to the digg submission. Suddenly a number of accounts (img2) begin to tweet the same link using different URL shortening services but often with the same ad text. The use of different URL shortening services in my opinion is an effort to confuse Digg to make visits appear natural instead of arising suspicion from the same link. (note: a small number of these pointed their links to Stumbleupon or Reddit)

How many ghost Twitter accounts spontaneously start promoting it?
@a_Technology, @BlissTribe, @cloudprav, @cybermotier, @darrylwilliamss, @designfront, @financial__news, @FollowInTheKnow, @gerryblack, @goodfinancial, @InvokeResults (same owner as @ReviveCoaching that also tweets it), @jc421, @juanarce, @Kafgemini, @lonnymiller, @KnowMoreTech, @kyrontrey, @mathewmurphy, @miketruong, @onlinedegr3e, @patelankur, @peterlubs, @redesocialesweb, @robertolorenzo, @robtesttweets, @Sandy3010, @sastiger, @spaceKebabji, @Srce_of_Errors, @SuperHuddle, @surveygizmo, @tchista, @tosumitgupta, @tracyoverton, @VictoriaNovikov, @voltronthecat (twitter account for a cat) @webvertise2, @wechartsdotcom, @weirdsht, @zombiegirrrrl

Unsettling Truths

Even though these accounts are crap, humans who follow these accounts might be able to generate the needed social media push. However how many of these 'gullible' people to have Digg accounts and how many of them use them on a regular basis? Furthermore can any of these accounts be tied to actual Digg votes from their administrators?

The unsettling thing is that I checked the people who dugg the article and account after account after account showed they were veteran users with many of them had Digg counts in the low thousands with some user approaching 100,000. For experienced users who have been on the site long enough to accumulate these numbers, how did they fall victim to something like this?

At the same time these repeaters are coming on line, Diggs start to magically appear and accumulate and the article is made popular. Is this all coincidence with one account influencing the votes of an entire community? Or could it be that this happens at around 2AM EST which ensures a lot of US Diggers are in bed and can't bury what amounts to an awful article (or are playing WoW or streaming movies). Furthermore another vocal segment, UK diggers, are commuting to work and settling in for the day.

The article currently has 526 views with 229 Diggs ... an incredibly high ratio considering the article's quality.

Something doesn't seem right.

Am I saying votes were purchased?

I'm not a cop and I don't want to assume anything but the votes on this article just don't seem to make any sense. The article is of incredibly poor quality, makes no conclusions, fails to argue any strong points, has no epic images and has some errors. Normally this kind of material gets 0 Diggs but somehow it racked up more votes than anything I've ever submitted to the service which includes some pretty awesome pics and epic criminal fails.

Considering the intelligence of the community and the general disdain for low quality material, I can only surmise that some of the established accounts that dugg the material didn't read it or perhaps received some considerations. I don't want to name names or speculate on what happened but I wouldn't be surprised if a pay-per-vote / submission service was behind it.

Comments (4) Trackbacks (0)
  1. Nice analysis.. but sour grapes.. If you know what I mean..

  2. Well this was submitted from an IP in India so I am assuming you are someone connected to this. Look you guys have found a way to exploit Digg but I don’t envy anyone who has to spam a web site in order to get traffic.

  3. Dear Will
    It would be useful to get your facts correct – ReviveCoaching and InvokeResults have no affiliation besides our connection on twitter – we are not the same company – a very brief look at our pages would make that clear. And @BlissTribe is not a ghost twitter account – again you would know that if you took more than 2 seconds to look at it – it is new and still building followers, but still happy to interact! And for the record I rarely use Digg, but that particular story caught my eye.

    Not a lot of credibility here dude – please research more thoroughly in future before you cast aspersions.

  4. Sorry Lisa. I’m sorry I didn’t sort the vomit spewed by both accounts to see who is throwing up wine and who is throwing up cheese. I have more credibility that someone whose profession is “life coaching”


Leave a comment

(required)

No trackbacks yet.

Bad Behavior has blocked 65 access attempts in the last 7 days.