Ad fraud is a multi-billion dollar a year problem and according to one estimate, by 2025, the digital ad market could be the second largest revenue source for organized crime behind drug trafficking.

Ad fraud comes in numerous forms. There's click fraud, for example, which for years has been a thorn in the side of marketers using pay-per-click advertising channels.

And in recent years, the rise of the programmatic has created a significant new ad fraud front. As Jeremy Hlavacek, VP of Global Automated Monetization at The Weather Company, explained, "Companies running the exchanges have perhaps been a little bit liberal in terms of who they let into that exchange."

The complexity and opaqueness that is often present in the programmatic ecosystem has led to inventory spoofing and unauthorized sellers, among other things, problems that the industry is now trying to stamp out with solutions like Ads.txt.

Driven in part by fraud concerns, marketers have been turning to alternative types of digital advertising, such as native ads and influencer marketing, which appear to be less vulnerable to fraud.

But are they really less vulnerable to fraud?

Influencer marketing agency Mediakix has sparked headlines by demonstrating just how easy it is to scam in the now billion-dollar influencer marketing business.

In a blog post, it explained how it created two fake Instagram accounts, one, @calibeachgirl310, using photos of a model obtained during a one-day photo shoot and the other, wanderinggirl, using stock photos. It then purchased followers for these accounts, at a cost of $3 to $8 per 1,000 followers.

calibeachgirl fake insta

Initially, Mediakix limited the number of followers it purchased to 1,000 followers per day because it was "concerned that purchasing too many followers at the onset would result in Instagram flagging the account" but it discovered that it was able to add up to 15,000 fake followers at once "without encountering any issues."

Finally, Mediakix paid around 12 cents per comment to generate fake comments on its accounts and between $4 and $9 per 1,000 likes to generate fake likes. For each photo on its fake Instagram account, the agency purchased between 500 and 2,500 likes and 10 to 50 comments.

The real fun began once the two fake accounts had 10,000 followers. "Once we hit this threshold, we were able to sign the accounts up for a wide range of [influencer marketing] platforms," Mediakix explained. And it started applying for opportunities on these platforms.

Ultimately, before Mediakix revealed its findings the fake influencer accounts were successfully able to land two paid brand deals for each account under which the non-existent influencers were "offered monetary compensation, free product, or both." 

For obvious reasons, Mediakix's experiment is raising eyebrows as it demonstrates that with modest effort and investment, it's possible to create out of thin air "influencers" who don't really exist and therefore aren't likely to influence anybody.

While this kind of fraud does not affect the upper echelons of the influencer marketing world, where high-profile celebrities like Kim Kardashian are said to rake in five and even six figures per sponsored post, the implications are increasingly important. As Mediakix explained:

Brands and advertisers eager to reach audiences on popular social media channels and seeking quick entry into the influencer marketing space, are turning to platforms, automation and micro-influencers in hopes of making the media buying process more turn-key and easier.

By automating their influencer marketing efforts and working with so-called micro-influencers, marketers risk falling victim to the kind of fraud that Mediakix has demonstrated is not just theoretically possible but can be successfully executed in the real world.

While this doesn't mean that marketers should avoid influencer marketing, or automation and micro-influencers, it is a reminder that the fraud risk is not limited to a few digital channels. Indeed, all channels are vulnerable.

Where the dollars flow, fraud will go.

As a result, marketers will need to be more vigilant about how they plan and execute their influencer marketing strategies. And just as many are starting to demand more of their vendors in other channels, such as programmatic, they would be wise to demand that the influencer marketing agencies and platforms they work with don't ignore this threat.

Patricio Robles

Published 16 August, 2017 by Patricio Robles

Patricio Robles is a tech reporter at Econsultancy. Follow him on Twitter.

2642 more posts from this author

You might be interested in

Comments (4)

paul kontonis

paul kontonis, CMO at WHOSAY

For our analysis at WHOSAY, influencer follower fraud is the highest at the micro-influencer level 0-50K level. As you go higher up in followers, the fraud levels drop substantially.. In addition, fraud is only an issue if you are relying on organic reach for your marketing campaigns and if you are, you are doing influencer marketing all wrong.

11 months ago

Mihnea Boiangiu

Mihnea Boiangiu, Online strategist at AdVision Interactive

That's a good article Patricio, but you missed something. How can marketers and agencies spot fake/fraud influencers?

@paul - At some point we purchased a shoutout on very popular account with over 1 mil followers on IG. The post got 6-7k likes, about 20 comments. Needless to say, the end results were lower than our expectations. Hence, I suspect them for artificially bumping up the numbers. The fraud could be higher on micro-influencers, but that's not the rule of thumb. For instance on micro-influencers collaborations you pay less and usually you get some photos from the deal. But what happens when you pay 3-4 figures for a shoutout? Is that something worth doing?

11 months ago

paul kontonis

paul kontonis, CMO at WHOSAY

@mihnea - we define 4 levels of influencers (the followers counts are not absolute but more of a guide): Micro-influencers (<100K followers), Influencers (100-750K), Trailblazers (500-2MM+), Icons (1-5MM+). On a platforms like Facebook and Instagram, organic reach is estimated between 2-10%. But talent alone does not impact performance, it has a lot to do with the creativity and quality of the content and distribution (some platforms are better than others for certain objectives). We base all our findings upon the 300+ campaigns we have run for top brands. Each campaign usually involves a few influencers and always includes paid promotion. We do not rely on organic reach only.

11 months ago

Patricio Robles

Patricio Robles, Tech Reporter at Econsultancy


That's a good question and as this problem becomes more widely recognized, I think we'll start to see best practices emerge.

When it comes to micro-influencers and agencies/platforms that allow brands to work with them in an automated or semi-automated fashion, I think brands will expect the agencies/platforms to do more thorough vetting of accounts.

Brands realistically are not going to go through lists of micro-influencers, checking to make sure the accounts are real. In a lot of cases they don't even know who the micro-influencers are because they rely on the agencies/platforms to select them.

The fake accounts that Mediakix created have some tell-tell signs of being fake, including short account age, stock photos, dubious-looking comments from dubious accounts, etc. The agencies/platforms probably would have noticed these things...if they were looking.

11 months ago

Save or Cancel

Enjoying this article?

Get more just like this, delivered to your inbox.

Keep up to date with the latest analysis, inspiration and learning from the Econsultancy blog with our free Digital Pulse newsletter. You will receive a hand-picked digest of the latest and greatest articles, as well as snippets of new market data, best practice guides and trends research.