Facebook Fraud

Facebook’s “legitimate” ad system for gaining page likes can still produce the same kind of fake-fan problem as outright click-farms—leading to inflated follower counts, suppressed engagement, and a pay-to-reach trap that benefits the platform.

The story begins with Virtual Bagel, a Facebook page that promised a joke business model: “we send you bagels via the Internet.” In 2012, BBC technology correspondent Rory Cellan-Jones created the page and bought likes to test what a like is actually worth. Two routes exist. One is the prohibited route: buying likes from sites such as BoostLikes.com, which rely on click-farms in countries including India, the Philippines, Nepal, Sri Lanka, Egypt, Indonesia, and Bangladesh. Workers there are reportedly paid about $1 per thousand clicks. The other route is the “allowed” route: paying Facebook for page promotion via ads.

Cellan-Jones paid $100 to Facebook and targeted the ad to the UK and the US—yet the likes poured in fastest from developing countries. Within a day, the page gained over 1,600 likes, mostly from places like Egypt, Indonesia, and the Philippines. The suspiciousness didn’t stop at geography. Many new followers looked like classic click-farm artifacts: for example, a Cairo-based account named Ahmed Ronaldo filled its profile almost entirely with Cristiano Ronaldo images and liked thousands of pages, while showing little real engagement with the Virtual Bagel page.

Facebook later reported deleting 83 million fake accounts in August 2012, about 9% of its total at the time, and some celebrities saw noticeable drops. But the deletion didn’t remove the underlying issue of fake or low-quality likes. The transcript’s author describes receiving Facebook emails offering $50 in free promotion for a modest page. After using the offer, likes tripled within days and kept rising by the thousands per day—eventually reaching around 70,000. Yet engagement on posts didn’t improve; it even declined.

The reason: fake likes behave differently from real followers. Facebook initially distributes each post to a small slice of page likes to gauge reactions. When those likes are fake or disengaged, Facebook sees low interaction (likes, comments, shares) and stops expanding distribution. That creates a paradox where follower counts rise while reach falls. The platform then earns twice: first from ad spend to acquire fans, and again when low engagement forces page owners to pay to promote posts.

The transcript adds a practical limitation: fake likes can’t be deleted in bulk; page owners can only target around them. Attempts to “solve” the problem by excluding click-farm-heavy countries also fail. A test with Virtual Cat—a deliberately awful page—still attracted likes quickly even when ads were targeted only to the US, Canada, Australia, and the UK. The accounts liking the page appeared to like enormous numbers of unrelated items (everything from major telecoms to random household products), suggesting that ad clicks and likes can be generated without genuine interest.

A proposed hypothesis ties it together: click-farms may click ads for free, then “launder” their activity by liking other pages so their behavior blends into normal-looking patterns. The transcript’s conclusion is blunt: Facebook has little incentive to remove fake likes at scale because doing so would expose ad revenue tied to non-genuine clicks and would undermine the engagement suppression that keeps advertisers paying to reach audiences that aren’t really there.