Op-Ed: Talking Influence WTF Series: WTF is a Bot?

August 26, 2021

This op-ed originally appeared in Talking Influence as part of the WTF Series.

In influencer marketing, there are no benevolent bots.

If you’re in the world of digital, you’re likely no stranger to the rampant issue of fraud. Fraudulent views, impressions, you name it – your campaigns have likely fallen victim to it at some point in time.

In influencer marketing, fraud has become an increasingly sophisticated challenge – 68% of marketers reported experiencing influencer fraud in 2020, and we know this number is on the rise. Why? As influencer marketing has seen exponential growth over the last several years – set to grow to $13.8 billion this year – bad actors have also been steadily taking hold, and have been deceiving smart brands, marketers, and influencers along the way. And in an industry where trust, authenticity, and transparency are so vital to the success of all parties (and really, the legitimacy of the entire ecosystem), bots are quite simply a problem that we cannot afford to ignore.

Not your friendly chatbot…

Flashback a few years ago and brands were dazzled by influencer accounts with massive followings and millions of likes. And unfortunately, some influencers took advantage, and purposefully manipulated the system by buying bots.

Today, we know the infatuation with follower count is slowly fading away, with quality, relevance, and a meaningful alignment of values increasingly becoming more important when brands and influencers enter into business partnerships. This being said, brands still want to know they’re reaching the audience that they’re paying for. 

When we talk about bots in influencer marketing, we’re not talking about the friendly chatbots that answer your questions and give you that scarily good styling advice on your favorite shopping site. We’re talking about the bots that massively inflate follower counts and leave marketers spending millions on campaigns that are meant to reach real people. And when your campaigns aren’t reaching as many real people as you thought they were, having any handle on true performance (the holy grail of influencer marketing today) is a real kick in the teeth. 

So, what are bots?

Ultimately, bots are followers that don’t exhibit human behavior – while today’s bots are pretty darn good at displaying very human-like behavior, there is a randomness to the ways that real humans behave that is nearly impossible to replicate on a large scale. 

Today, there are three main types of bots that dominate the social media sphere. And remember folks, there are no good bots in this story!

  • Mass follower bots: these are accounts that follow 7,500 users (the max allowed). The profiles will contain posts that are either stolen from other users or random stock photos. Luckily, this type of bot is quickly being eliminated by platforms like Instagram.
  • Hacked accounts: these are the most “cost-efficient” bots and are exactly what they sound like – real accounts that are hacked, that then unfollow everyone the former account holder was following and start following people at random until it reaches a set number. When that number is reached, all activity is ceased in the account.
  • Cheap bots: while hacked accounts usually cost around a dollar, cheap bots go for a thousandth of a cent. They are nearly universal in their behavior – no posts, no followers, and no activity other than following other users. In most cases, once they follow 500 people, they will stop all activity and the account is essentially abandoned.

Now knowing the different types of bots, it’s clear that not all bots are created equal when it comes to their sophistication and ability to deceive. But we also know that as digital evolves, bots will too, and so marketers need to get real about the problem. 

Don’t ignore the problem 

Luckily for the industry, there is smarter, better, more advanced technology out there that is helping the industry get ahead of the issue of influencer fraud and bots. But, like most problems, acknowledgment is key – marketers simply can no longer stick their heads in the sand, especially when influencer budgets are growing at a scale and pace they are today. 

To protect and encourage even more growth in the industry, vetting influencers and their audiences for bots is a crucial and necessary part of the process. So, what’s the secret behind some of the best technology out there today? The key is to look beyond the surface – meaning, looking beyond an influencer’s immediate followers. 

Sophisticated technology is leveraging adaptive machine learning to closely analyse behavior at the follower level, the followers’ followers level, and even a level deeper at the followers’ followers’ followers. Going this deep on the audience helps identify all of the fraudulent (but dangerously human-like) behaviors of bot accounts that we addressed above, like level of activity and engagement. These AI-driven technologies are evolving in real-time to bot behavior, which makes them essential to the future success and growth of influencer marketing.

If you’re not yet convinced, think about this for a second…

Imagine you’re spending $1,000 per post on an influencer with two million followers…and you find out that out of those two million followers, only one million are actually real humans. Probably wouldn’t make you feel too good about that $1,000 a post price tag, right? If you knew this information upfront before agreeing on rates, you could negotiate better based on real human followers, and spend those incremental marketing dollars elsewhere, like on another influencer with one million followers. 

Let’s be clear, bot detection is not about finding a way to pay influencers less; it’s about accessing more value out of influencer marketing and building more trust across all stakeholders involved. 

So, hear this: before you invest your marketing dollars in that seemingly perfect influencer partnership, do yourself a favor and make sure you’re getting the full picture.