Social Media Advertising 6 min read

Facebook's Fake
Like Economy

Substantial evidence suggests Facebook's advertising revenue is propped up by fraudulent engagement. Here's what the experiments proved — and why the platform has little incentive to fix it.

Facebook Fake Likes

Facebook's advertising model has been under scrutiny for over a decade. The core problem is simple: a Like is only valuable if it represents a real person with genuine interest. The evidence suggests that, at scale, it often does not.

The disconnect between Likes and actual engagement was first noticed around 2012, when analysts observed that accounts responsible for 80% of page Likes accounted for just 1% of user engagement. That gap is not a rounding error. It is a structural problem.

Two Ways to Buy a Like

There are two ways to purchase Facebook Likes. The first is through a third-party service — what Facebook publicly labels a scam. The second is through Facebook's own advertising platform. Facebook's objection to the former is that purchased Likes come from accounts with no genuine interest in your page. The uncomfortable truth is that Facebook's own ad programs produce the same result.

The US State Department made this problem visible at scale. After paying $630,000 to acquire 2 million fans, the department found that engagement sat at just 2%. The numbers did not add up — because the Likes were not real.

80%
of Facebook Likes came from accounts with only 1% engagement rate
$630K
spent by the US State Department for 2 million fans with 2% engagement
83M
fake accounts removed by Facebook in 2012 — without deleting the Likes they left behind

The Experiments That Proved It

Several journalists and creators ran controlled tests to understand exactly what Facebook advertising was delivering. Two experiments stand out for their rigor and their results.

Virtual Bagel Ltd. — BBC, 2012

BBC Technology Correspondent Rory Cellan-Jones created a fake Facebook page for a fictional company offering bagels you could download and eat. He spent $10 on Facebook advertising. Within 24 hours, the page had 1,600 Likes. It ultimately reached over 4,000. Analysis of those accounts revealed clear patterns consistent with click farms — yet he had not hired one. He had used Facebook Ads. Facebook removed 83 million fake accounts that same year but left the fraudulent Likes intact.

Read the BBC article →

Virtual Cat — Veritasium, 2014

Science YouTube channel Veritasium created a page called Virtual Cat, explicitly describing itself as "intentionally blank and meaningless" and asking anyone who liked it to explain why. The creator targeted the ad only to cat lovers in the US, Canada, Australia, and the UK — specifically to exclude countries where click farms typically operate. After spending $25, the page had 262 Likes. Eight people saw the pinned post asking why they had liked it. Zero engaged. Every account that liked the page had liked thousands of others.

Watch the Veritasium video →

Why Click Farms Use Facebook Ads

"Click farms click Facebook advertisements for free — it keeps them off Facebook's fraud radar, and Facebook collects the ad revenue either way."

Click farms in developing countries — predominantly in India, the Philippines, Nepal, Sri Lanka, Egypt, Indonesia, and Bangladesh — pay workers roughly $1 per 1,000 clicks. Rather than liking pages directly, which triggers fraud detection, they click on Facebook advertisements. This looks like legitimate ad traffic to Facebook's systems. Facebook charges the advertiser. The click farm avoids detection. The advertiser receives a worthless Like.

The incentive structure is the core issue. Facebook profits from ad clicks regardless of whether those clicks represent genuine interest. Resolving the problem thoroughly would reduce revenue. The platform has not resolved it.

How the Fraud Loop Works
01
Advertiser Pays Facebook

A business runs a campaign to grow its Facebook page audience, paying per click or per Like.

02
Click Farm Sees the Ad

Farm workers encounter the ad through their normal feed activity and click it — appearing as organic traffic to Facebook's detection systems.

03
Like Is Recorded

The page receives a Like from an account with no interest in the product or service. Engagement remains near zero.

04
Facebook Keeps the Revenue

The ad budget is spent. Facebook has been paid. The advertiser has a larger audience that will never buy anything.

The Bottom Line

The evidence is not new, and it is not ambiguous. Controlled experiments conducted independently across multiple years and geographies reached the same conclusion: Facebook advertising at scale generates significant volumes of fraudulent engagement. The platform has the data to know this. It has not solved it.

For businesses allocating marketing budgets, the question is not whether fake Likes exist. It is whether your campaign's engagement metrics reflect real customers — and whether the platform you are paying has any structural reason to ensure that they do.

A Like with no engagement is not an audience. It is a vanity metric with a price tag attached.

Sources: BBC News — Virtual Bagel experiment, 2012 · Veritasium — Virtual Cat experiment, 2014 · The Next Web — Likes or Lies, Jaron Schneider, 2014