Facebook claims a third more users in the US than people who exist
Are we looking at an advertising house of cards?
Facebook promises advertisers access to more US customers than actually exist.
That's according to an investment analyst who had long held that Facebook is misleading the market on what its actual digital reach is, and recommends a "sell" on the social media giant's stock.
Facebook has an extensive and sophisticated ad-buying system that assures potential advertisers it can reach no fewer than 41 million of a core target group of 18 to 24-year-olds in the United States.
The only problem, analyst Brian Wieser of Pivotal Research Group pointed out in a note to customers, is that there are only 31 million of them that actually exist in the US, according to the official census data. The same gap in reality also holds for other groups, including the next most-targeted group of 25 to 34-year-olds.
There are very few independent checks on the accuracy of information and data provided by Facebook, Google and other digital companies that monetize their huge user bases by offering targeted advertising.
Facebook is also renowned in the ad-buying business for being a black hole, particularly when it comes to things like paying for Facebook to get Facebook users to "like" your Facebook page – a circular money flow that boggles the mind.
But for those hoping to target millennials, the company's ubiquity is a good solution to a hard-to-reach group that does not use traditional media such as TV or radio anywhere near as much. Claiming more users than actually exist is a stretch even by Facebook's standard, however.
Of course, Zuckerberg's Monster has an answer for why this isn't a problem. The audiences it offers to advertisers are only estimates that do not match real-world figures "by design" according to the company, because they account for a "number of factors, including Facebook user behaviors, user demographics, and location data from devices."
Say that again?
The estimates "are designed to estimate how many people in a given area are eligible to see an ad a business might run," the company said in a statement. "They are not designed to match population or census estimates."
What does that mean in plain English? One of two things. Either Facebook knows better than the US government who is actually living in the country (tens of millions of illegal immigrants and tourists maybe?), or the website is flooded with fake profiles. We're willing to bet on the latter.
Despite recent often-criticized efforts to get its users to supply their real personal data, Facebook is still almost entirely reliant on users to self-report their age, sex, location and other information. There is also nothing to stop anyone from creating fake personas or profiles. So the question is: how many profiles are fake? It could be as high as half of them.
As has been noted repeatedly in recent years, on social media you may be spending much of your time and your advertising money talking to computer-generated bots that exist solely to game systems and boost specific posts or articles. As we saw with the fake news controversy earlier this year.
The problem is that Facebook has a massive financial incentive to present the highest possible figures. While Facebook does not charge per estimated audience but by user view, every ad view and click equates to real dollars.
It is a fair bet that when a company claims it has 32 per cent more users than actual live people, a big chunk of your advertising dollars is going down straight into the platform's pockets with nothing in return.
How is Facebook able to get away with it? Because advertising is a notoriously low-hit-rate business. For a banner ad for example, a 2 per cent click-through rate (CTR) – meaning two out of every 100 people that see it click on it – would be considered extremely successful. Most CTRs these days average about a quarter of a percent.
However, if Facebook's own figures are wildly off, it could lead to a big re-evaluation by ad-buying companies, and that would have an immediate impact on the company's revenues, profits and stock price.
Facebook is well aware of the problem. It has repeatedly been forced to apologize to advertisers for over-counting its own ad impressions (weird that it never seems to undercount).
In November last year, it noted that the daily and monthly stats it gave about visitors to people's Facebook homepages didn't account for repeat visitors – which is a huge deal for advertisers. It also admitted over-reporting "by 7-8 per cent on average" the number of people who had read articles on the site – for more than a year.
Lies, damn lies and statistics
That comes on top of admitting that its average viewing time stats for videos were over-reported by 60-80 per cent because it simply removed views of less than three seconds from that calculation while retaining the number of people who had viewed the ad.
In response to those debacles, the company launched a new Measurement FYI blog to act as "a new channel for regular information on metrics enhancements."
From Facebook's perspective, its efforts to get more solid user numbers are repeatedly met with objections from privacy advocates and lawmakers. It actually beat an effort by the Belgian government to stop it from tracking non-Facebook users, but the company is constantly being criticized for its efforts to reach into people's lives as a way of gathering ever more data in order to both validate what it thinks it knows about its users, and to come up with new ways to monetize that information.
The company also plays loose and fast with its privacy policies and tracking efforts. It continues to swear blind that it does not use its users' location data without their permission – and yet people continue to report being offered up "friend" recommendations when they have no connection to the person apart from that fact that they live close by. ®