Influencer fraud is everywhere, but only the data-rich social networks can stamp it out

Until the Facebooks of the world wake up and realise the long-term damage influencer fraud is doing to their precious social networks, it will always exist in one form or another – argues Nik Speller of the agency Influencer

“The key to improving the situation is three-fold: cleaning up the influencer ecosystem by removing misleading engagement; making brands and influencers more aware of the use of dishonest practices; and improving transparency from social platforms to help brands measure impact. We need to take urgent action now to rebuild trust before it’s gone forever.”

No sooner had those words left Keith Weed’s mouth, back in June, than you could hear the cries of anguish in offices up and down the land as bigwig marketing chiefs, at brands and agencies alike, threw down their copy of Campaign magazine and asked everyone and anyone just what they were doing about this new thing called influencer fraud.

The answer, of course, was likely to be muted. Everyone who works anywhere near the coal face in this industry knows about fraud and sees it on an almost daily basis; but, it’s oh so easy to turn a blind eye, bury your head in the sand, or come up with some sort of garbled nonsense of a justification.

“Follow / unfollow isn’t cheating — it’s just fast-tracking your way to an audience.”

“They said they only did it because their agent told them to.”

“How else are they supposed to get the swipe-up function?”

“Everyone in the industry does it!”

(These are all genuine messages I’ve received on the subject)

Especially when the incentive in this game is to get numbers, numbers, and more numbers; never really to prove impact of any sort. Of course, these excuses never washed; but, now influencer fraud has (finally) hit the headlines they most certainly don’t. And, it’s likely that every agency (claiming to be an expert in this industry) and any brand (whose invested significant resources in this industry) is triple-checking everyone they ever worked with and everything they’ve ever said, just to make sure how badly they’ve been burned or how implicated in fraud they’ve become.

For those of us who’ve been in this industry for a while now and have been banging the drum about influencer fraud for what seems like forever, it’s refreshing to see that (finally) others are talking about it – and with a loud voice too (even if it is slightly disappointing to see them claim to be ‘the first’; but, I guess that’s just marketing for you).

But, let’s get one thing straight here: no matter what’s being said about whatever super systems now exist, influencer fraud is far from over  – and it would be incredibly naive and irresponsible to claim that it is.

Smoothing the curve

The issue with fraud – unfortunately – is that it’s never straightforward, it never exists in just one form, and it never sits still and lets you discover it, without evolving. Fraud is an attempt to pass-off something, someone, or a behaviour as something, someone, or a behaviour that it actually isn’t. And, in an unregulated space, with profit to be made, businesses (which influencers are, both real and fake) will look to exploit that lack of regulation and push the boundaries as far as they can to make as much money as possible.

As regulations come in or the ability to detect fraud improves, sure, some careless Charlies out there might be immediately caught out for using the most obvious methods of cheating; but, the rest of the bad guys will just evolve what they do to avoid detection. Any system, no matter how advanced, that looks for an irregular pattern of behaviour to detect fraud will always be susceptible to those fraudsters who can mimic what is defined as a regular pattern of behaviour.

If your system looks for sudden peaks in followers, sudden explosions in likes, or large, unexplained steps in whatever other metrics you can measure, the fraudsters just look to smooth their curve, to something nice, neat and organic looking. As someone over in the States told me a few months back, bought followers in the United States used to be easy to identify by looking at an influencer’s demographics; if 90% of them are from Brazil, it’s highly likely that they’ve been bought. Once this was recognised as a weakness in the fraudsters approach, they simply switched to bought followers from the US –  and problem solved.

So, the system for detecting fraud can never be an immediate silver bullet ,  it has to evolve, continuously, to keep up with the changing approach of the fraudsters; and, that evolution is almost always going to be a step or two behind, as its the hunted, rather than the hunters, who are the ones with the biggest incentive to twist, cheat, and fool the rest of the industry. Even more so, if the rest of industry is only just starting to wake up and realise the extent of problem . A problem that it helped cause.

Friendly fire

Maybe the new systems for detection are doing this, though. Maybe they’re up-to-speed, running at pace with the fraudsters or even ahead of them; the people on-board, so in tune with the fraudsters that they can guess their every move and improve their system accordingly to always detect true, organic engagement versus the attempts of the frauds.

There’s an issue here too, for a cynic like me (and others I’ve chatted to on the subject). The smooth curve of organic engagement or audience growth doesn’t always apply and, where it doesn’t, the influencer in question isn’t (necessarily) a cheat. Viral content, by its very definition, doesn’t follow a nice smooth line; it explodes in ways that could suggest someone was up to no good.

I can think of plenty examples where I’ve seen someone’s follower growth and thought something was up,  only to be corrected by someone who knew that person better, pointing out that the content or account had done something that attracted insanely high engagement. The simple solution to this, where content is concerned, is to look at an average, over time, and see if this is a common occurrence; and, where audience numbers are concerned, do some digging and see why (and how) that account could have grown astronomically.

Again, though, life isn’t that simple. Viral content doesn’t always follow the same pattern (a video of mine on YouTube lay dormant for a month, before a few big folks shared it and the views rocketed up) and this lack of consistency in content performance opens the door for friendly fire, where those people whose virtues are being extolled in the big wide world outside of social media or who just produce incredibly good and ‘on trend’ content get shot down for cheating, unfairly.

Judge, jury and executioner

Friendly fire raises another important point in the fight against influencer fraud. If there’s to be one system to rule them all and be the (self-declared) arbiter of who is (and who isn’t) committing fraud, there has to be due process  – there has to be a defence to the prosecution and the right to prove or disprove the accusation. And here is one of my biggest discomforts with the hunt for the fraudsters. This week I’ve already heard (twice) the term ‘found to’ . As in, ‘influencers who have been found to be cheating will be x, y, or z’.

There is no ‘found to’ — there’s merely ‘suspected of’ and ‘accused’. No one out there (save the social networks) have a hold of the definitive proof of fraud . And all of us who’ve been playing this cat and mouse game for a while know that (aside from those careless Charlies) uncovering fraud is a hugely complex, deeply difficult task, that involves studying Social Blade graphs, buying tokens for HypeAuditor, reading the tea leaves, asking the questions, checking the notifications, putting out the feelers, saving the screenshots, crossing our fingers and praying our (educated) assessment is the right one.

There can be no ‘found to’. There can only be ‘suspected of’ and ‘accused’; and I cannot say that enough. If one system, tool, or agency is going to determine who is the cheat and who is for real, this opens the door to all kinds of friendly fire incidents, reputational damage for perfectly innocent people, and even (dare I suggest it) an incentive to buy likes, followers, and cheat on behalf of influencers you hate, in order to have them hung drawn and quartered by a system that allows them no retort.

Hiding in plain sight

There’s one final problem with a system which claims to be determining the bad from the good and, in doing so, becomes the arbiter of truth in this industry. With the bad influencers getting a black mark,  a Homer Simpson style red stamp across their knuckles which says cheat,  and the good influencer being waved through to lucrative brand deals – safely stamped with a halo on their crown – what of those of the former who do ‘smooth the curve’ and become recognised as the latter?

Those fraudsters who hide in plain sight, don’t get caught, and (consequently) get labelled as the good guys, are now (presumably) seen for once and evermore as legitimate influencers. They’ve passed the test and they have the seal of approval for brands to work with them.

But, what’s the process going forward? Are they tested again, in the future, to check they haven’t since turned to the dark side? Is their past revisited, in the future, as the ability of the system to detect fraud improves?

Or is that halo applied once and never revisited? Again, it can’t be said with anywhere near 100% reliability that these influencers have been ‘found to’ not cheat; it’s merely that – to the best of the system’s ability –  they’re ‘not suspected’ of having cheated. The language employed when dealing with fraud has to be careful, in both the positive and the negative. Undoubtedly, cheats will slip through the net and to label them, then, there and forever more, as clean-living influencers who haven’t touched a bot in their life is (again) both naive and irresponsible.

The system, whatever it is, has to evolve to match changes in –  and a better understanding of –  fraudulent behaviour. In doing so, each and every person who sails through with flying customers has to be revisited, scrutinised, and declared ‘safe’ to work with (for now); otherwise, any and all cheats who slip the net make a mockery of the system and their existence (even if it’s just one person) casts extreme doubt of any positive (and, perhaps, negative) conclusions from the system.

In conclusion

Tools, systems, tea leaves, and whatever else can be conjured up to tell ‘good’ from ‘bad’ in the influencer industry has to be celebrated and commended. But, as with anything that aims to clean-up a swamp (and, not to get all Trump for a moment, this industry is firmly built on a swamp), has to undergo the same rigorous scrutiny (if not more) as that which it aims to tackle.

No one system can (or will) ever be infallible. Fraudulent influencers will continue to exist, all the while both the incentive and tools to commit fraud exist. Sure, as the systems improve, many of the crudest fraudsters will get caught (but, so will many who are innocent); but, the savvier hawks out there will twist, turn, adapt, and change to get through the net and continue reaping the rewards of their approach.

This is why, as Keith said, we need the social networks to act. They’re the only ones with access to the full data, to tell us who is using automated systems, who has thousands of inactive followers, whose hours of activity are well outside the norm. Until the Facebooks of the world wake up and realise the long-term damage influencer fraud is doing to their precious social networks, it will always exist in one form or another.

And those of us who care about this industry and want to see it survive will have to keep battling against them, try to uncover their tactics, and make sure we don’t incentive and reward the cheats; rather that simply declaring that one silver bullet has ended the war.

Speller has spelt out the dangers of influencer fraud

Nik Speller is head of campaigns at the Influencer agency and is based in London, in the United Kingdom – a version of this article was first published on Medium


Get the latest media and marketing industry news (and views) direct to your inbox.

Sign up to the free Mumbrella Asia newsletter now.



Sign up to our free daily update to get the latest in media and marketing