Customer engagement is a goal for virtually any business with an online presence (i.e. virtually any business these days). So it follows that businesses want followers, likes and the like. But what happens when the people following you on social media aren’t real? And just how widespread is the phenomenon of fake followers?
- 11.2% of Facebooks 1.23 billion monthly users are fake (Facebook).
- 14.2% of large US retailers Twitter followers are fake (Entrepreneur.com).
- Google is now periodically auditing YouTube views to verify that they’re genuine (Google).
Fake Followers Can Lead to Real Results
It’s usually easier for something that is already popular to gain greater popularity than something that is completely unknown. Nobody wants to eat in an empty restaurant after all. Hence, the utility of fake followers: businesses can use an initial pool of fake followers to give the impression of popularity to those (real) people that they want to engage. But does this actually work?
In his Medium article ‘(Fake) Friends with (Real) Benefits’, data scientist Gilad Lotan provides an account of his investigation into the potential fruits of buying fake Twitter followers. For the low price of five dollars, Lotan bought 4000 followers for a bot account he created that goes by the handle @gilgul. And while these followers were fake, they all had profile photos and bios of such level that a passing glance at them wouldn’t immediately reveal anything suspicious. Upon a closer inspection, it wouldn’t be difficult to discover their spurious origin.
Almost as soon as the 4000 fake Twitter followers were obtained, @gilgul’s Klout score skyrocketed. Moreover, because Klout works with Bing, @gilgul also enjoyed higher rankings in Bing’s SERPs.
An increase in the @gilgul account’s real followers was observed as well. And while the number of fake followers began to decline because Twitter regularly deletes such accounts, over the next few months the number of real followers continued to grow. Litan speculates that the considerable number of fake followers made the @gilgul account look more credible to real people and that this perceived credibility made these people more inclined to follow @gilgul.
To sum it up, Litan’s study suggests that buying fake followers can have real benefits insofar as perceived popularity can lead to real popularity.
Fake Followers Aren’t Condoned By Social Media Companies
None of the major social media companies condone fake profiles or likes. Facebook, for instance, is now actively combating fake likes and names on its platform. This means that if you do buy fake likes or followers–even just to jumpstart your presence–your base of fake followers may very well disappear before you have a chance to build your real following as fake profiles are routinely deleted by Facebook, Twitter and other platforms.
Fake Followers Don’t Engage and Don’t Convert
If the goal of your social media marketing is to engage people whom you want to convert into leads, it will be difficult to do so with people who don’t actually exist. And when it comes to lead generation, social media is a tool few businesses can afford to ignore. Indeed, Business2Community reports that 54% of B2B marketers have produced leads using social media, with 40% of those leads generating revenue. The same report also notes that 49% of marketers report social media as the most difficult lead generation method to execute, but the benefits make social media well worth the effort.
There is evidence to suggest that fake followers can help you build a real audience that you can engage. However, fake followers will not themselves, of course, improve engagement or lead generation. And while fake followers may help build a real audience, you must weigh this with the risk of contradicting the policies of social media companies as well as the ethical issue of deceiving your (real) audience. Furthermore, if your goal is to improve engagement with users, solutions called recommender systems or recommendation engines are helping media publishers, ad agencies and even the social networks themselves produce positive, dramatic results.
A recommendation engine is a system that, as its name suggests, generates recommendations for users. For instance, a VOD site may use a recommendation engine to suggest videos to users which are similar to those which they recently viewed. However, not all recommendation engines are equally effective.
A good recommendation engine uses a host of techniques and algorithms to analyze user behaviour and deliver accurate item recommendations, which usually results in improved engagement. In contrast, rudimentary systems often suggest items simply on the basis of shared properties. For example, suppose a person is watching a video titled ‘How to Cook Chicken with Christopher Walken’. The latter type of recommender system–suggesting items only on the basis of item properties such as titles–might suggest a video titled ‘How to Hunt Chickens with Christopher’. The words used to describe these two videos, though similar, refer to two very different videos–and probably different audiences as well. The more advanced recommendation engines are able to determine that users watching the ‘How to Cook Chicken with Christopher Walken’ are probably more interested in cooking than hunting.