Click fraud is now officially rampant.
In a recent report, Bloomberg Business Week cited a study which said that fake traffic will cost web advertisers $6.3bn this year.
And in some ways, we shouldn’t be surprised.
Advertising on the web offers huge opportunities for click fraud, and if there is an opportunity to get ‘free’ money, then we should expect nefarious individuals to exploit it to a great extent.
But social media is supposed to be different. The social platforms control the distribution of their content, including ads, so we can rest assured that our fans, followers, and (most importantly) ad targets on social media are real people, right?
There is a lot of concern that many profiles on social media are fake and so the impressions and clicks we are buying are as worthless as a 1-pixel banner ad.
How do we know?
Social media fraud isn’t widely reported, but there are a few clues to the extent of the problem.
- Audience numbers don’t add up. In another post, I estimated Facebook is over-reporting its reach in Singapore by 22%. A lot of the extra profiles could be fake.
- The platforms admit it. Facebook’s most recent annual report revealed that ‘undesirable’ (spam) accounts exist and claimed that they are up to 2% of monthly active users.
- Investigative reporters write about it. The Guardian did an investigation into a click farm in Dhaka, Bangladesh in 2013 and more recently in 2015 Doug Bock Clark wrote a fascinating report on one in the Philippines for New Republic.
- You can buy ‘likes’ yourself. Search Google for Facebook Likes, Twitter followers, or LinkedIn connections and you will see an active marketplace willing to create fake social profiles for a surprisingly small amount of money.
Each of these corroborate with each other to some extent, so it’s almost certain that businesses exist to create fake social profiles on a mass scale.
It’s worse in Asia
And there is some evidence that the problem is worse in Asian markets.
No location-based data has been released by the social platforms but Apsalar, a mobile app tracking solution, reports that Hong Kong, India, and Indonesia have the highest incidence of false clicks versus Germany, France, and Italy with the lowest.
And Grace Liau, Asia-Pacific GM for Vivaki, a digital ad firm, says ”markets like Thailand and India have one-tenth of the actual click-through rate [CTR] that they do ‘on paper’.”
Why fake social profiles are a problem for brands
Even though the fake profiles are, in some ways, harmless, the people who activate them are tasked with keeping them ‘alive’ and away from the occasional cull by the social networks.
To do so, the fake profile users will have to log in occasionally, view ads, click on them, like items, etc. all of which cause problems for brands who are advertising on the platform.
The most obvious problem for brands is that they are spending money on impressions and clicks that go nowhere. It’s marketing money down the drain that would be better invested elsewhere.
But potentially an even bigger problem is that fake social profiles interfere with metrics.
Brand marketers are constantly asked to deliver ROI reports on a market-by-market basis. And if they are trying to measure success in a market flooded with bots, these metrics will be totally off.
Sure, you will get the views and the clicks but the conversions will be far below expectations, and the meagre results on social media could stop digital advertising or prevent launching in the market.
So, what to do about it?
There are a few things that brand marketers can do to alleviate the effects of fake social profiles.
1. Use performance-based metrics.
One way to know for sure that you are not over-delivering to fake profiles is to use metrics which are based on the performance of your brand.
Facebook, for example, offers ad pricing on Website Conversions or on Lead Generation.
2. Carefully segment your target audience.
When you buy media on social platforms you are given a wide range of options through which you can segment your audience.
If you do so into meaningful groups, it can be quite obvious when one is bot-ridden and performing poorly and another is a more reasonable benchmark.
3. Use your CRM for targeting.
Facebook, Twitter, and Google allow you to upload your CRM and target based on actual users.
You can also build look-a-like audiences from your email list which may not be entirely bot-free, but will almost certainly be better than interest or affiliation targeting.
Sadly, though, fake profiles seem to be ubiquitous now and the numbers will probably grow as long as people keep purchasing Likes, fans and followers – and we shouldn’t expect that to change soon.
But if you start with performance-based metrics, carefully segment your target audience, and use your CRM to help refine your audience further, then you can avoid the worst effects of fake social media profiles.