According to recent estimates, Instagram has about 95 million bot accounts, which represents roughly 10% of all accounts on the platform. Distinguishing genuine influence from deceit can be daunting. With our fake follower audit at Influencer Hero, we empower brands to detect fake followers with precision. By leveraging sophisticated algorithms and machine learning, Influencer Hero scrutinizes engagement patterns, follower authenticity, and activity metrics to identify fraudulent accounts. This article explains in detail how we detect fake followers and how you can use this data to optimise your influencer marketing ROI.
Fake followers are artificial or inactive accounts created to inflate an individual’s follower count on social media. Often generated by bots or purchased through dubious services, these accounts lack genuine engagement and interaction. These accounts are often created solely for the purpose of inflating audience size or other illegal activities. This is typically not a person who consumes content through the account.
As fake followers are not real consumers which could potentially become paying customers you would like to make sure to avoid paying for influencers with large amounts of fake followers. If an influencer has bought fake followers it is also an indicator of unauthentic behaviour which is probably not the type of influencer you would like to associate your brand with.
To be more specific we conducted one of the largest research on this topic to date. Based on a data set of 10,325 commercial influencer posts our data scientists computed the correlation of various metrics to conversion. Based on this research we found that audience credibility aka fake followers have a strong correlation with conversion performance highlighting the importance to analyse this metric.
To visualise the results of our study better we categorised the 10,325 influencers into the following 3 segments.
In the graph below you can see that high performing influencers on average only had 21.5% fake followers, followers by medium performing influencers with 24.7% and low performing influencers with 27.4%.
There are two reasons influencers have fake followers:
All influencers have some fake followers. This is simply due to the nature of social media platforms which attract bot accounts for fraudulent behaviour. These accounts are often used for various purposes and might randomly follow accounts to mimic fake human behaviour.
In most cases if an influencer has an above average percentage of fake followers they might have bought these followers to inflate their audience size. Influencers sometimes do this to look more attractive for brand partnerships.
There can be many different reasons why fake followers / bot accounts exist:
There are service providers which make money by selling fake followers. They often create fake accounts in mass which they then monetise in different ways. One way is to sell the following to influencers or other people who want to artificially inflate their followers size or engagement.
Data providers which deliver insights to companies often use bot accounts for data extraction from social media platforms like Instagram. The tools used to crawl data are often referred to as web spiders or web crawlers.
Fake followers / bot accounts are also often used for fishing and other fraudulent activities. Accounts might be used to exploit old people by deceiving them by acting as another person to extract sensitive information or login details. Bot accounts are also frequently used for spam to mass outreach people to sell products or services in an illegitimate way.
At Influencer Hero we use the most sophisticated predictive models in combination with Ai and machine learning to detect fake followers. The machine learning model we use is based on an academic model released by Phds from Switzerland. The model takes various data points into accounts through which we can identify fake followers with 98% accuracy. Some of these data points which are taken into account are:
With Influencer Hero we detect between different types of audiences so you know exactly how the audience of an influencer is structured to improve your campaign effectiveness.
These are straight up bot accounts. Fake followers which you want to avoid at all cost. In the example below you can see the audience details of David Beckham. In this case the fake follower % (suspicious accounts) is 14% which is healthy. As previously mentioned all influencers have a certain % fake followers. We typically recommend to avoid influencers with a higher fake follower % than 30% as this could be an indicator of having purchase fake followers.
Mass followers are accounts which follow over 1,500 profiles. The reason we highlight this is because it is very difficult to reach accounts which follow a large number of profiles. The accounts are in essence competing for limited inventory / display space. Meaning that followers which follow >1,500 are very unlikely to see the post of an influencer. Sometimes this also indicates engagement pots where accounts follow each other to artificially increase engagement and following. Although these aren’t bot accounts it might be good to keep an eye on this metric when deciding to work with influencers.
These are accounts which have over 10K followers indicating that these are influencers. In the example below David Beckham seems to have 5% influencer followers.
These are real people, consumers and potential buyers for your brand. The higher this number the better. In the example below David Beckham has a healthy real people following of 63%.
Similar to mass followers in the previous section we have detected the follower numbers of the followers of the influencers. This might sound a bit double but to analyse the audience effectively you want to understand how many accounts the followers of the influencers are following. The higher this number the lower the chance that they see the sponsored content of the influencer you are planning to work with. In th example of David Beckham 49% of his followers follow less than 500 accounts, 10% follow between 500 - 1,000 accounts, 23% follow 1,000 - 1,500 accounts and 19% follow more than 1,500 accounts.
As previously mentioned we typically recommend being a bit more cautious if an influencer has over 30% suspicious accounts. Now this doesn’t mean that you shouldn’t work with this influencer as there are a lot more things to analyse to make this decision. If an account has over 50% suspicious accounts we typically recommend not to work with that influencers. Other data points which are important to analyse and showed a correlation with conversion in our research:
To access the full research report click here.
There are various free tools which let you analyse fake followers some of which are:
Data provided by Influencer Hero’s fake follower tool:
Data provided by influencer hero’s fake follower tool:
We have a detailed review of Upfluence fake follower tool to compare and check the best fake follower tool.
Data provided by Modash’s fake follower tool:
We have a detailed review of Modash fake follower checker and also 9 other fake follower tools to review and select the best tool to detect fake followers.
Data provided by Influencer Hero’s free influencer audit Chrome extension
In summary, detecting fake followers is crucial for optimizing influencer marketing campaigns and ensuring genuine engagement. Influencer Hero leverages advanced machine learning algorithms to scrutinize engagement patterns, follower authenticity, and activity metrics, providing brands with precise insights into their influencer partnerships. By distinguishing real followers from bots, Influencer Hero helps companies make informed decisions, drive authentic connections, and enhance their ROI. As social media continues to evolve, tools like Influencer Hero are essential for maintaining credibility and fostering genuine online communities.
Schedule a Demo with one of our media experts below.
.png)
.png)
.png)