Fake Fans

Digital Footprint

In recent weeks a few people have asked me about the current state of social media. They asked if fake fans were still a thing. The short answer is YES. The long answer is far more interesting.

There are many reasons why people get involved in social media and blogging. It’s a sloppy argument to suggest that it’s all about “free meals”. Yes that happens, mostly with those who are guilty of thinking small and not realising the value of what they do. Some people become bloggers as a creative outlet, and over time become very good at it. Some decide to take a leap and turn their social media into a platform for their career.

In this context fake fans become especially interesting. If you sell exposure on your account, or sell expertise on the basis of your account, then purchasing fake fans or fake likes is a massive deception.

I can’t recall how many times I’ve been asked the question, “How can I tell if someone is buying fans?”

Whenever I see an Instagram account with awful photos and lots of followers I am immediately suspicious. Photos don’t have to be dreadful to peak my suspicion, they can be technically competent but lacking in soul and creativity. Unless you are knee deep in social media platforms yourself it can be hard to tell if someone is being honest with their numbers.

But sometimes people leave a breadcrumb to follow. One such case popped up recently.

The Instagram account was brought up by someone at an agency, as part of our conversation. They handed me their phone to show me the account, and at first it looked impressive. Big numbers. A reasonably consistent style but lacking in nuance, character, personality and inspiration. I had my laptop open and pulled up the web version of Instagram and scanned across their recent posts.

This is where it gets interesting. The first thing we noticed as we browsed the account was a massive lack of comments. Each post was scoring around 1500 likes, but the comments were between 2 and 10. This is easier to see on the web interface, the numbers pop up as you scroll over and you can pick out patterns more clearly.

We started clicking on a few of the posts to see what kind of comments they were getting. It became clear that even when an image tallied 10 comments, half of those were actually from the posting account anyway. They add one comment to load up hashtags, and then respond to every comment that comes in. I compared that to my own Instagram account, which has roughly half the number of followers, and historically I get comments ranging from 10 to 100. I get a wide range. Sometimes it’s 50, or 20, or next to nothing.

I have a few turkeys in my feed it seems.

Then I noticed something even more interesting. A lot of the comments from one photo to the next were being made by the same accounts. These comments used a familiar combination of emojis, or even just a single emoji. It was as if the same person was logging into different accounts and throwing a few lazy comments to help boost some seriously pitiful engagement.

So we looked a little closer at those accounts that were frequently generous with their comments, and found something even more curious. That same photographic style was obvious on the ‘commenting accounts’. Lacking in nuance, character, personality and inspiration — but the same style of photos none the less. Another of the commenting accounts appears to post a photo no more than once a month, and yet comments routinely of the main Instagram account.

This particular Instagrammer looked to be managing a handful of client accounts, and was using those accounts to make their own look more loved. With just a few minutes of digging we got a pretty good feeling for how dodgy this blogger is. At this stage I will call them a “Fake Influencer”.

In a former career I was a research scientist, and I learned a handful of useful statistical tools that help to sort feelings from facts. Spotting a pattern in a data set is one thing, proving it is another. I wont go into the gritty details, because numbers can be boring where a simple graph can be convincing. Below are two ‘digital footprints’. The first one based on a genuine Instagram account that is populated with genuine fans, likes and comments. The second is based on our fake influencer.

Natural Digital Footprint
Dodgy Digital Footprint
Same chart with 1200 likes subtracted from each post

We ran one more comparison with this influencer’s account. If we subtract 1200 likes from the same data for each post, we get a more natural looking spread of engagement. In other words, the data suggests that they simply buy a fixed number of likes for each post which is swamping the organic engagement they would otherwise have.

Consistency is a funny thing. As a publisher of content (photos, words, articles, blogs) you need to be consistent with your work. It’s the bedrock of a successful business. Your fans however are not consistent. You can post a great photo one day and find next to nobody agrees with your aesthetic, and then post something the next day that you feel is boring only to find it gets regrammed 17 times and for the next 12 months it keeps popping up for another round of regrams just when you thought it was dead.

The internet is made of people and people are never ever consistent. That’s what our ‘digital footprint’ graphs are trying to express, the degree of ‘consistency’ in feed engagement. In the natural world we talk of “bell curves”, where most of the population exists in the middle and some of it exists on the fringes. The shape of the curve tells you something about the variation in a population. The same is true of engagement on Instagram.

My own Instagram account
My Own Digital Footprint

So what does that orange chart tell us about the engagement of our fake influencer? It suggests very very strongly that they are buying likes for their account, but not comments. It also suggests they are not very good at hiding dodgy efforts.

That account was easy to spot for fake likes and faux comments. They made a few mistakes and were not very smart in covering their tracks. Other bloggers are a little smarter and I wish there was a statistically based analysis tool available to scrutinise Instagram accounts.

I submitted the fake influencer account to a few online tools that rate and validate the ‘engagement’ of Instagram accounts, and they completely failed to spot the fake likes. That was disappointing. Maybe the commercial tools are a little more nuanced when looking for insta fakes, but I don’t have the budget to test them. On the flipside, one account that I know has been buying fans (instead of likes) managed to get past these same tools and only got outed for a 10% fake fan base. So I just don’t trust these tools for spotting certain kinds of fakery.

When I tested a few tools that claim to be able to spot fake followers, I found something equally problematic — they can report false positives. I submitted my own account to one of these tools as a baseline, and it gave a wildly inaccurate figure. It was using a sample size of just 150 (out of 42,000) and then listed a few of the “most fake” accounts that it detected. I had a look at the first few it reported as fake followers and concluded they have a very bad algorithm. They were real people, not fakes.

I dug a little deeper and came across FameAudit. Again I used my own account as a baseline for testing and it found some accounts that definitely look fake to me. This site claims that the average Instagram account is being followed by an average of 10% fake accounts. These are bot accounts set up to spam, like and trade as real followers. It turns out we all have fake followers whether we buy them or not.

So I dropped in the account that we had already determined to be buying fake likes. FameAudit reported that 47% of their followers were not genuine. That’s a massive figure, and in keeping with what we already suspected.

I would dearly love to name and shame this particular fake influencer. Outing them would be an act of malice however, and nothing good comes from malice. I wont even hint at their gender or city, because if you’re reading this you may well know their account. I can tell you they are the kind of person that gives blogging a bad name. They are rude to their hosts, oblivious to their companions and exhibit a wide range of ways to treat other people like garbage. It seems they are also treating their professional clients and PR contacts like garbage too.

While their influence is fake, the dollars being spent by other people as a consequence of their deception is real. How would you feel if you found your social media manager was a minimum of 47% fake? That expenditure won’t do much to help your business, but it might give a little boost to a fake influencer’s profile and bank balance.

Sudden Spikes in Followers?