How brands can spot a fake influencer

Outgoing Unilever CMO Keith Weed recently alerted brands on fraudulent influencers and said, “The dark side of influencer marketing is that people are buying followers. The worst is having bots retweeting and doing views.”

But talking about the positive side of influencer marketing, Weed said the market is too good to let a few bad apples affect it completely and we need to take some steps to ensure it is fit for future growth.

So, how can a brand weed out those bad apples?

Although brands are employing influencers across platforms, Instagram is beating other social platforms hands down as engagement rates on the platform are 15 times higher than on Facebook and 20 times higher on Twitter, according to a global industry study report.

With the growth of automated follower services, fake followers and generally inorganic accounts, the need for authenticating social media accounts has become profound.

There are plenty of bot detection tools (free and paid) that claim to offer a solution through varying methods of reliability, including analysing follower patterns, engagement behaviour and mapping bot farms to track anomalous behaviour. However, every algorithm-based result thrown by tools has its flipside. Hence, working with a professional who has experience in spotting fake influencers will drastically increase your chances of spotting them.

Spotting follower spikes

Big or erratic spikes in followers may mean dubious techniques, including purchased ones. However, legitimate media mentions may also see rapid follower growth. Similarly, no spikes don’t always mean no fake followers.

Location mapping

If an influencer has a high following of followers in countries that aren’t the influencer’s home territory, it looks suspicious. For example, if an Indian influencer is creating local content, yet has high volumes of followers in countries such as Brazil, Turkey or China, it is likely these are fraudulent, especially given there are substantial numbers of fake-follower operations within these territories. This factor is very easy to surface within the social audience demographic analysis.

Engagement

An influencer with huge numbers of followers with only a handful of interactions (likes, shares, comments) could indicate fake followers. If an influencer has a very low engagement rate, particularly given the specific platform’s average, this should raise a red flag.

However, it’s not enough to take low engagement as the only deciding factor, as fraudulent profiles can appear to have healthy engagement if bots capable of engagement via likes and comments are also being used. Check comments for specifics. Comments like ‘great picture’ and ‘love your style’ probably means they are fake.

Take away

Just because an influencer has some fake followers, doesn’t mean they have paid/asked for them. Equally, bot comments are usually implemented by the account, commenting as a way to draw traffic to their page.

To find potential potholes and check fake influencers, BuzzInContent.com tried a few tools below:

HypeAuditor

HypeAuditor enables you to perform an audit on Instagram accounts. It can detect fake followers, engagements and audience authenticity by using an advanced AI fraud detection system.

In HypeAuditor, you can check comment authenticity, inorganic spikes in followers as well as engagement rates to test the legitimacy of an Instagram influencer. Like Fake Check, the pricing model is on a per search basis. The audience quality score and follower quality analytics are highlights of this fraud detection system.

SparkToro

SparkToro doesn’t help you get rid of your fake followers; you’ll have to spend money to do that. It’s just a tool to analyse the followers of any Twitter account. It looks at over 25 factors correlated with spam/bot/ accounts to return a percentage of followers that are thought to be fake.

Upfluence

The Upfluence plugin is a (free) extension of our software that allows anyone to see the analytics of any social profile online, including the number of fake followers.