Dark Nodes and Dodgy Connections; Dealing With Fake Followers
Posted by Brian Kelly (UK Web Focus) on 26 August 2012
“It’s About Nodes and Connections”
In a recent post I described how Social Media is About Nodes and Connections and explained “the importance [of] the network effect, with a growth in the number of nodes (the bloggers, the contributors, the Twitter users) leading to a growth in the number of connections (the posts, the comments, the tweets, the retweets) which help in the development of new insights and new ideas“.
But whilst many users of social media, including those working in higher education, are making use of such network effects to support their professional activities in legitimate and ethical ways others are seeking to exploit network effects in ways which may be considered unethical.
Fake and Inactive Connections
An article on such approaches was published on Sunday 26 August 2012 in the Observer. The article asked How many Twitter followers do they really have? and explained that although Lady Gaga has almost 30 million followers on her Twitter account only 29% of these are “good”. The remainder are either fake or inactive accounts. Whilst inactive accounts will be simply those used by people who are lurkers or who may no longer have an interest in using Twitter, a fake account is set up to follow people or send out spam.
The article describes that there is now a market for the sale of Twitter followers. ”One kind of software identifies Twitter accounts that include keywords such as football, and “follows” these accounts in the hope they will reciprocate. Other programmes create artificial accounts and sell them by the thousand. On the Fiverr website, 2,000 followers can be bought for $5“. As can be seen and shown in the accompanying image there are clearly several providers of such services.
The Observer article described StatusPeople.com, “a British start-up company has pledged to root out and expose the phantom, fake and fraudulent followers being used to massage the numbers claimed by celebrities, politicians and the merely insecure within the Twittersphere“. Although at pricing ranging from £25 to £100 per month I can’t imagine there will be many subscribers, I’m pleased that we are seeing public awareness of spam problems and solutions being developed, starting with such auditing tools.
In a research context we are seeing how Scholars Seek Better Ways to Track Impact Online who recognise that “research that used to take months or years to reach readers can now find them almost instantly via blogs and Twitter”. However as Ernesto Priego pointed out on the Guardian’s Higher Education Network’s blog a few days ago for Alt.metrics “‘quality of engagement matters as much as retweets”.
I hope that pressures to maximise ‘impact’ will not lead to researchers buying Twitter followers in the hope that this will increase the numbers of downloads of their papers. But if they do I suspect that this will not be productive – I suspect that we will see developments to alt.metrics tools which will help to identify fake followers. We might also see the development of alt.metrics measures which will provide more sophisticated measures which give high weighting to active engagement on Twitter rather than passive consumption.
What’s a Twitter User To Do?
Clearly one should not buy followers! But what should you do if you see an influx of people who have started to follow you?
Such information is not easily found. On a mobile device the Twitter client now enables you to see not only messages sent to you but also recent actions, which includes tweets you have sent which may have been retweeted or favourited, as well as people who have started following you.
A few nights ago nine people started following me, as illustrated. I have not got into the habit on the bus on the way to work of block the obvious spam followers – in this case @NetEquityLoans and @DrinkTampico.
In order to decide whether Gregg Thorpe was a spam follower or not a trivial amount of further investigation was needed: the Twitter ID @1stplaceranking gave the game away, as did the suspicious Twitter statistics – 2 tweets, 4,681 followers with 5,124 users being followed!
Loreen Deeka also appeared to be a spam account and checking recent tweets confirmed this.
These four accounts were subsequently blocked. As well as applying such approaches regularly I have also used software such as the SocialBro desktop application which gives me a profile of my Twitter environment including suggestions of suspicion followers.
The Observer article made me appreciate that there will be a need for professional users of Twitter, whether individual of corporate, to have a policy on how they deal with misuse of Twitter. I’ve therefore decided to document my current policy (which is liable to change):
- I will monitor new followers and block obvious spam followers and other followers which I feel are inappropriate.
- I will use Twitter auditing tools periodically to identify inappropriate followers and block them.
I suspect there will be a need for such policies for institutional Twitter accounts. Are people aware of any which have been published? I’m also interested in the approaches which individuals may take in blocking followers. Or do people chose to take no action?
Twitter conversation from Topsy: [View]