The simple truth about social networks is that we use them predominantly for saying something to someone who is hopefully not only listening, but will also find what we say interesting enough to want to share it with their other friends/followers/fans. In the Twitter context, Twitterers want to have followers who are real, attentive and responsive. The default preference is human followers.
You can also read: Are social network relationships about quantity or quality?
In this post, which was inspired by an update in my Digi-DIY Tips page where I posted about a tool that was recommended to me by a friend, I test the extent to which Twitter followers of 25 influential Twitterers are fakes.
Loosely defined, fake followers are fictitious Twitter accounts or just bots attached to downloaded apps, and that can actually send malware as genuine Twitter messages on your behalf. I receive these all the time, and have become vigilant after my Twitter handle was corrupted by one such malware.
I have put the 25 Twitterers included in my analysis into 5 different groups: (a) my pick of 5 South Africa’s influential non-media Twitterers; (b) my pick of 5 South Africa’s influential media Twitterers; (c) my pick of South Africa’s 5 influential politicians on Twitter; (d) South Africa’s top 5 celeb Twitterers; and (e) Twitter’s top 5. All the stats in the graphs below were drawn on March 20th, 2013.
A. Faker scores of my pick of 5 South Africa’s influential non-media Twitterers
While my selection of Twitterers in this group was somewhat random, I picked people who Twtrland refers to as Power Users. Backgrounds of the selected people vary widely, and it includes a banker (Michael Jordaan), a business strategy consultant (David Graham), a brand and reputation architect (Thebe Ikalafeng), an Internet entrepreneur (Rob Stokes) and an infopreneur (myself, of course (•‿•)).
Analysis of fake scores indicates that this group of influential Twitterers is followed by mainly humans who are also active on Twitter, as indicated by the high GOOD score.
B. Faker scores of my pick of 5 South Africa’s influential media Twitterers
The selected media Twitterers in this group come from TV, radio and print. I have also chosen to include Rob Byrne, the SABC‘s main traffic reporter, to add a dimension to the group.
About half of this group’s followers are active humans. Ashraf Garda, the media personality who is on both radio and TV and follows social media developments avidly, has the healthiest profile of followers. While the number of fake followers is not too big in this group (at 16%), INACTIVE Twitterers make up more than a third of the average total number of followers.
C. Faker scores of my pick of South Africa’s influential politicians on Twitter
I chose the most well-known politicians for this group, including the country’s first citizen – President Jacob Zuma. Malusi Gigaba is in the ANC NEC and he is the current Minister of Public Enterprises. Helen Zille is leader of main opposition – the DA, and her right-hand woman – Lindi Mazibuko – is the party’s Parliamentary Leader. Zwelinzima Vavi, the General Secretary of Cosatu, would probably argue that he is a trade unionist and not a politician. However, the two are closely intertwined in Mzansi. Thus, I am comfortable with him being part of this group.
Notice that contribution of fakers in this group is almost directly correlated to total number of followers in this group. On average, a quarter of the group’s followers are fake.
D. Faker scores of South Africa’s top 5 celeb Twitterers
The list of South Africa’s top 5 celebs was extracted from Twitter Counter.
One in every 3 followers in South Africa’s top 5 celebs is a fake. In addition, Bonang Matheba has the least number of active human followers. As can be seen, the lady is popular with INACTIVE Twitterers.
E. Faker scores of Twitter’s top 5
The list of top 5 Twitterers can be found on Twitter Counter.
On average, close to half of top 5’s Twitter followers are fake. What is even interesting is that Justin Bieber, the top Twitterer, has only 13% of active followers, and a whopping 67% are fake. Maybe Justin is a super-human, or a robot for that matter! *Just asking*
Comparison of follower profiles across the selected 5 Twitterer groups
Comparison of follower profiles of the 5 Twitterer groups I selected for this post reveals interesting insights.
Here are my 2 key insights from the comparative follower profile graph:
- It appears that follower quality deteriorates as number of followers grows. Intuitively, this makes sense. The only worrying factor is the magnitude of deterioration, as shown by the RED line. Good question would be: Is there a tipping point in the accumulation of followers? Even if we knew, would it matter?
You can also read: Twitter followers and Facebook fans: 2 buy or not 2 buy?
- It appears that the likelihood of strongest inverse relationship depicted in the graph above is between Fakers and Good followers.
Time for a disclaimer. The insights above are to be treated only as indicative. Stating them as fact needs prior statistical validation.
Three questions on a parting shot:
- Should influential Twitterers worry about fakers as their follower numbers grow? If this is warranted, is there a solution to this challenge?
- Does it make sense to buy followers? If yes, how can we verify that what we are buying is not fake followers?
- Should brands stop buying promotion tweets of influential Twitterers such as Kim Kardashian, given that they could be throwing good money after bad?
All graphs in this post were generated using infogr.am.