Detecting Fake Influencers: Everything about this fraud

Influencer marketing has grown into a powerful marketing tool that lets brands reach thousands of potential customers. With the right influencer, a brand can secure high returns on their investments. And there’s a very good reason behind why this happens. 

People are more likely to purchase a product or service that’s been recommended to them by a family member, a friend, or someone they look up to. For a long time, magazines, billboards, TV ads and other forms of advertisements were how brands reached a large pool of potential customers. But the rise of social media brought people closer together in more ways than one. 

These days, you’re more likely to generate and convert leads by investing in a social media influencer. People now have the freedom to choose the content they want to engage with. This includes people on social media who they may not know or never meet. But through building these small online communities, people who follow certain influencers become interested in how that person lives and what they have to say. 

Influencer marketing capitalizes on this relationship between a social media user and their followers. An influencer is paid a small fee in return for marketing a certain brand to their followers. The resulting sales made from that exchange can be far greater than through many traditional forms of advertising. That’s why many brands now consider influencer marketing to be a core part of their marketing and sales strategy. 

However, every coin has a flip side. Influencer marketing has the potential to help brands grow and expand their consumer base. It’s a very lucrative business for influencers, with top influencers earning hundreds of thousands of dollars from a single post. This prosperity has also attracted scammers, fraudsters, and people looking to make a quick buck as well. Influencer fraud is very much a real part of influencer marketing, even in 2021. It’s more than likely here to stay, and that’s why knowing how to detect fake influencers is crucial to saving money and building sales figures.

What is Influencer Fraud?

Not every ‘influencer’ you’ll find on social media is the real deal. A person is considered to be an influencer when they have a large number of followers. A micro-influencer has anywhere between 10,000 – 50,000 followers. A top influencer can have millions of followers. 

For small and large brands, this is great news. If one influencer is paid a small fee to promote a certain product or service, then that influencer can immediately reach thousands of people with their promotion. Since the people following the influencer already trust them and have built a rapport with them, they are more likely to purchase the product or service. Right? 

This only applies to social media accounts that have real followers. Influencer fraud happens when a person, a bot, or a group of people create the illusion of being an influencer, without actually having any real influence on the platform or among its users. In 2019,  influencer fraud was projected to cost brands nearly $1.3 billion.  For 2020, that number went up to $1.6 billion. 

Research conducted by Cheq, together with the University of Baltimore, revealed that the drive to prove that a social media account has a vast reach led to the birth of new fraudulent businesses. These exist for the sole purpose of creating and selling fake followers, creating fake engagements etc. for fake influencers.

Referencing a study from Paquet-Coulson, the Cheq research showed that people can ‘buy’ up to 1000 followers for YouTube for just $49. For Facebook, 1000 followers cost $34. The rates were even less for platforms like Instagram ($16) and Twitter ($15). Given that Instagram is the most commonly used influencer marketing platform, that’s bad news for businesses and brands who don’t do their background research. 

Twitter, Instagram and other social media platforms immediately got to work to detect and remove fake accounts. In fact, the number of fake accounts on different social media platforms were so great that Twitter themselves lost over seven million followers from their primary Twitter handle. 

Instagram also conducted their own research to identify and analyze data related to influencer fraud. They found that influencer fraud is actually on the rise again. Once news of fake accounts with fake followers and low engagement rates became public knowledge, brands were understandably concerned.

Instagram had introduced a measure where users who used third party applications couldn’t like or comment on posts on the platform. Initially, the rate of fake influencer engagement dropped by up to 1.7%. This especially applied to accounts that had fake and unauthentic followers. However, within a few months, fake engagement again rose to about 1.2% on unauthentic profiles. This indicates that there are people out there who are actively trying to circumvent any restrictions that social media platforms place their way. 

What are the Different Kinds of Fake Influencers?

There are three main kinds of influencer fraud that brands and businesses encounter most often: 

• Automated Bots: These are software designed to like, share and follow other accounts. Their primary purpose is to boost and inflate influencer metrics to give the appearance of authenticity to the account. However, the engagement these bots provide is not real engagement. 

• Buying Followers: An Instagram profile with 100,000 followers but around one or two likes per post is an example of influencer fraud. There are businesses out there who provide fake followers, likes and comments in exchange for a small amount of money. Accounts that use bought followers don’t ever have a good engagement ratio. This implies that people are not engaging with such accounts. Brands that market through such fake profiles will never be able to secure their ROI or witness any boost or growth in sales figures. 
  
  
• Pod Groups: Pods are groups of real people, sometimes numbering thirty, often more, that give each other likes, shares and follows, to inflate each others’ influencer metrics. They generally join a group, such as a Facebook group or an Instagram group chat. There they help each other grow their accounts without actually knowing each other or engaging with each others content in any meaningful way. Hashtags such as #likeforlike or #followforfollow are examples of the kind of hashtags pod members use to find each other. 

It’s crucial for brands to work with real influencers if they hope to find business and sales success through influencer marketing. 

Examples of fake accounts and fake engagements

These accounts showcase how bots create fake profiles. Bots are designed to generate large numbers of profiles – almost as many as 50,000 if not more. Once created, these accounts then ‘follow’ each other. There are some tell-tale signs that can immediately inform you whether a profile is real or created by a bot.

Here, @sta_azz showcases what influencer fraud looks like perfectly. With absolutely no posts, they’ve ‘gathered’ 1301 followers. They are following 595 accounts – indicating that pods little influence here. Since bots can generate thousands of fake accounts which they then automatically follow, most of the account following @sta.azz are possibly bots. 

Note how their name is a mix of alphanumeric characters. Bots tend to create accounts that combine random alphabets, adding a few numbers at the end of the alphabets in the username. These accounts have no pictures at all or just a few ones – another warning sign. 

This wouldn’t be unusual, unless you take into account that the commenters are leaving behind the smileys on multiple posts. This indicates the most likely use of pods to generate engagement. 

How Can You Detect Fake Influencers?

In 2019 alone, brands invested around $1.9 billion in influencer marketing in the United States and Canada alone. Of this, nearly $1.4 billion went towards influencers on Instagram. Almost one-fourth of this amount, roughly $255 million, unfortunately went to fake influencers.

Influencer fraud has turned into a major risk factor for brands and businesses that want to promote their products or services through influencers. Despite this, influencer marketing continues to be a popular way to reach new audiences and boost sales. The middle ground here lies in finding authentic influencers who have a real following and can truly get their audience to invest in the products and services they promote. 

There are two ways to detect fake followers: 

1. Manually Conducting Background Research on an Influencer’s Profile

This involves: 

• Screening profiles to look for red flags. Some common red flags that indicate influencer fraud include profiles with no display pictures, or stock images that have nothing to do with the account. 

• Follows from the account as well as followers to the account have nothing in common. This is done simply to inflate the follower count, with most followers being fake followers. 

• Bots as followers, or following bots. Bots are automated software that follow, leave likes or unrelated comments on posts. A popular example of this is leaving an ad in the comments section of a post, or commenting something irrelevant ‘nice pic’ on a picture of a sky or a hamburger. 

• Usernames that look like they’ve been randomly generated. These can include numbers, be alphanumeric, or otherwise have nothing to do with the actual content in the profile. 

• Using stock images, especially copyright free stock images, as posts. This is most often seen on platforms like Instagram, that rely on visuals. 

• The ‘influencer’ is based in an area of the world that has nothing to do with their followers. An account primarily has followers based in the United Kingdom, but the ‘influencer’ is listed as being from Russia or Saudi Arabia etc., is almost always a fake account. 

• There are other influencer metrics that can be used to gauge whether the influencer is real or fake. Checking their background history to see how long they’ve been posting, looking at their followers to engagement ratio, and more can help brands further isolate fake Influencer accounts. 
  
  

2. Artificial Intelligence and Data Driven Tech Based Solutions  

Many brands simply don’t have the time to sift through data on various influencers to differentiate between real and fake influencers. However, this also increases the chances of engaging with fake influencers who don’t have the reach or ability to provide proper ROI. This is where AI and other tech based solutions come into play.

• AI and machine learning together can look through vast amounts of data from thousands of influencer profiles to separates the real from the fake. 

• Users are identified and their Instagram statistics analyzed. 

• Influencity’s AI-solution, for example, utilizes a metric called ‘Followers Quality’. Here, variables such as the number of followers, accounts being followed, bio description, profile picture and more are analyzed to determine the authenticity of users. Real influencers with real audiences can easily score over 80 on this metric. 
  
  

Selecting the right influencer to market through can help brands budget their marketing strategies accordingly. ROI’s only become a reality when they real influencer’s promote brands to their followers.

Take a look into the following influencers, they seem to have a good community behind: @missgutsstein, @magarbuglio, @diaashechka, @djjerrydavila, @donomar_11

Would you be able to say if any of these influencers have some type of influencer fraud on their profiles? We will analyze these profiles later using Influencity and see what happens. 

How Can You Tell if You’re Interacting With a Fake Influencer Account?

Influencer fraud is more common than people realize, even if the extent of that fraud isn’t always as great. According to a recent study that investigated more than 5000 influencer accounts indicates that:

• Typically, the average influencer will purchase nearly 23% of their followers. 

• 30% of influencers have been found to generate flags when it comes to buying engagements. 

• Nearly 90% of influencers fail to reach and engage with less than half their total audience. 

• More than 50% of influencers fail to reach and engage with less than 25% of their audience. 

• Influencer engagement metrics can be off by 55%. 

• Impressions created by influencers can be off as well, by about 48%. 

In this digital landscape, how can businesses learn how to identify and deal with fake followers and influencers?

Measuring Real Influencer Metrics Against Fake Accounts

Brands and businesses that want to avoid falling prey to influencer marketing, should watch out for the following: 

1. Rapid Increase in Followers

A real influencer needs to take time and invest their resources in order to grow their social media presence. This typically takes a few years. Going viral or a sudden spike in popularity can boost how fast the influencer’s follower count grows. But this isn’t the standard for everyone. 

Most people spend time to grow their presence, establish relationships with their followers, and then reach a place where they can be called an influencer. Fake accounts indulging in influencer fraud don’t need to establish relationships or organically grow their accounts. All they need is the illusion of having influence over actually being influential. Accounts that were recently created but that boast a high number of followers are always fake. The number of followers may even be added over time in an effort to showcase ‘authenticity’. But the pace at which followers are added are still much faster than the time it takes for a real influencer to establish themselves. 

2. Engagement and Impressions vs Follower Count

An authentic influencer builds their relationships through interacting with their followers. This is seen through likes, comments, tags, shares etc., depending on the social media platform being used. Not every follower will react or engage with every post from an influencer. However, the ratio between their follower count and engagement on their account reflects their authenticity. A person with 1000 followers on their Instagram account can easily expect around 50 engagements at least per post. And these engagements usually have something to do with the post in question. 

Fake followers on the other hand, hardly ever interact with the accounts they’re following. These usually tend to be bots or pods. The interactions they leave behind tend to be of two types:

• Bots will either not engage, or leave behind suspicious comments that lead to unrelated advertisements. The advertisements they link to can lead to malicious software, viruses, or sites that can harm a user’s PC, tablet or smartphone. 
  

• Pods leave behind generic comments. These comments will say things like ‘nice’, ‘good’, ‘like it’ or other generic phrases that don’t speak to the post itself. A quick look at the account the comment was left from will show that the account they posted to left behind a similar comment on their posts. 
  
  

Bots and pods are used to inflate influencer metrics in order to boost asking prices from brands for promotions and endorsements. 

3. Follower Demographics

Brands focus on certain influencers in order to reach their target audience. If a brand wants to sell streetwear to teenagers, then they’ll want to approach an influencer who can promote to that age group. For most influencers, this isn’t an issue since their followers will reflect their own demographic and interests. 

Fake followers are instantly visible simply by browsing through the profile’s list of followers. There is no coherence between the kind of followers that account is following, or is being followed by. This is an instant giveaway. These fake accounts may be devoid of profile pictures, not have content on their profiles, and leave behind little to no likes or comments. 

4. Platform Specific Red Flags

Some red flags are understood in the context of the platform being used. For Instagram, this relates to pictures, videos, stories and other visual elements. A fake account will not invest in quality photoshoots, maintain an aesthetic feed, or use the right hashtags. Using stock images or images off from Google are tell-tale signs of influencer fraud on Instagram. 

For YouTube, the red flags are seen through the view count, the likes to dislikes ratio, as well as the comments left behind. A video with a high number of views but few or no comments is likely a participant of influencer fraud. This applies to various social media platforms – from Facebook to Twitter to TikTok. 

How Can You Avoid Falling for Influencer Fraud?

The  Influencer Marketing Benchmark Report 2020 revealed among other data, the following: 

• Influencer fraud has been increasing. 
• Nearly 2/3rds of all respondents to their survey admitted to interacting with or experiencing influencer fraud. 
• Nearly 91% of respondents considered influencer marketing to be an effective way to boost sales and reach their target demographic. 
  
  

The report detailed that in 2020, influencer marketing was set to grow to nearly $9.7 billion. But, influencer fraud has been a part of the influencer marketing industry in recent years. This has created more awareness among digital marketers and brands. The criteria for selecting the right influencer is no longer limited to their follower counts and reach. Now, the rate of engagements is considered to be more important. At the same time, the focus has been shifting from top influencers to micro-influencers. 

Micro-influencers tend to have a smaller reach, usually no greater than 50,000 followers. This limited reach also provides a greater potential for building strong relationships with a real audience. The ROI when it comes to hiring micro-influencers can be greater than hiring top influencers or celebrities. 

But, digital marketers and brands still have to do their own background research to determine the quality and veracity of the influencer they’re hiring. This can take time and effort. Brands and marketers have to look through thousands of accounts and conduct specialized research each time they want to hire an influencer. Is there a better way to do influencer marketing? Yes there is. 

Why AI and Tech Based Influencer Fraud Solutions?

As an alternative to physically conducting research that could span hours, tech based solutions could be the answer to influencer fraud. Influencer technology automatically goes through hundreds and thousands of accounts and locates real influencers over fake ones. 

People engaged with influencer fraud are always striving to conceal their deceit. It may be difficult for a person to know which account is real and which isn’t.

Keeping up with influencer fraud is just as difficult as locating it. Those engaged with fraudulent behavior are constantly developing new ways to convince brands and marketers that they’re the real deal. 

Let’s look at the examples we were talking earlier: 

Meet @Missgutsstein.

@Missgutsstein looks like she’s well on the way to becoming a popular influencer. She’s already on the top edge of the micro-influencer category. Since micro-influencers are considered worthwhile investments currently, doesn’t that make her a good choice? 

Not if her statistics are to be believed. AI based software like that from Influencity makes use of influencer statistics and metrics such as followers quality and audience overlapping to determine whether an account is real or not. 

Let’s analyze her followers deeply:

@Missgutstein has received a followers quality of 26.48%. Authentic accounts receive a score of 80% at least. This automatically indicates that this account is related to influencer fraud. Looking at the data, she has only 1.84% engagement. That means that a minuscule portion of her nearly 50,000 followers engage with her account.

More than 50% of her followers, nearly 35.82K, are listed as ‘doubtful followers’. This implies that this number could come from bots, pods or paid followers who will not buy any product she promotes or endorses. From brands and businesses, this account is not worth investing in. 

Looking further at her ‘average activity split’, she’s received less than 1K interactions on her profile. For an influencer with over 50K followers, this is an abysmal engagement ratio. Brands and businesses that invest in her are likely to make low, if any, returns on any investment if they hire her as their chosen influencer. 

Check here the full analysis of this profile.

Let’s see the other profiles and check if they have a good followers quality or not:

@magarbuglio has a 29% followers quality, very low actually. Other indicators like Engagement shows you that this profile has fake followers. Check the full analysis here.

@djjerrydavila has only 26.k real followers, although his account shows more than 140k. In this case, the engagement is also a red flag. Check the full analysis here.

@diaashechka with over 400,000 followers seems a good influencer profile. But understanding her community, we see that only 38% of those followers are truly real. Check the full analysis here.

@donomar_11 also looks like a good profile, however his followers quality is the lowest in this list. Check the full analysis here.

AI and tech based solutions harness the power of machine learning to grow with the data that’s presented to them. Unlike people who may struggle to go through thousands of profiles, tech based solutions can do the same in a matter of minutes. And as the data they’re analyzing grows and reflects new trends in influencer fraud, the influencer technology automatically grows with the data. This means that tech based solutions can keep pace with changes to and new kinds of influencer fraud that keep emerging in the industry. 

Tech based solutions allow brands and businesses to quickly locate the right influencer for them, as well as separate fake influencers from the real ones. Brands, by looking at a profile’s earned media influencer statistics, can also better prepare their marketing budget. The right blend of a powerful influencer and the right marketing campaign can bring in real returns on investments. They can generate leads, boost sales, and enable brands to convert their target demographic in a more efficient way.  

Want to analyze any influencer profile, sign up here and try it for free for 7 days.