Transfin.
HomeNewsGuidesReadsPodcastsTRANSFIN. EOD
  1. News
  2. Explained

What is Ad Fraud? How Does It Work?

Editor, TRANSFIN.
Nov 28, 2020 12:45 AM 4 min read
Editorial

Last week, news reports emerged that Facebook was offering advertisers tens of millions of dollars in credits over a glitch in the tool it used to inform advertisers about the efficacy of their ad campaigns.

What Facebook Did

This glitch reportedly skewed data and overestimated how well an ad campaign was performing, thereby driving up advertisers’ ad spending even though the alleged app downloads, website clicks or product purchases they were paying for may have never happened.

For its part, Facebook said it fixed the glitch in September and was “working with impacted advertisers” to ensure fair compensation. However, this isn’t even the first time the company has been at the centre of an ad revenue-related controversy.

In 2016, Facebook was taken to court over a miscalculation in how it calculated average video views, which resulted in a false inflation in views by 60-80% for years. The Menlo Park company was accused of knowing about the disparity for a year but conveniently failing to disclose the same.

Facebook’s missteps and intentions aside, this episode does bring to light one of the biggest threats to the c. $200bn online advertising industry - ad fraud.

 

What is Ad Fraud?

As the name suggests, ad fraud involves misrepresenting impressions, clicks and/or conversion data in online advertisements. This is done to generate more revenue by charging advertisers for exposure and clicks that never took place. And this is usually executed through fake audiences, fake users, fake clicks or fake traffic.

FYI: Simply, speaking, the online advertising industry involves clients or advertisers (companies or individuals wishing to sell their products, promote an event or ask people to download their apps) outsourcing their digital ad spending plans to digital marketing agencies, who tie up with ad companies (like Facebook and Google) to display these ads.

Ad fraud a big deal. It cheats advertisers of their money and distorts data and marketing budgets. It is estimated to have cost advertisers $35bn so far in 2020. But because ad fraud is so hard to detect, the real number is probably much higher.

 

Who are these Ad Fraudsters?

Broadly speaking, there are two categories of ad fraudsters.

One involves individual or professional hackers. They set up fake websites where they display ads. Then bots are used to drive up page traffic artificially and create bloated ad impressions - on which the hackers make high margin profits. (Fun fact: Nearly half of all web traffic is driven by bots, not humans. And not all these bots are “bad”.)

The second category is more sophisticated. This involves an elaborate and alleged nexus between some digital marketing agencies and ad networks wherein they sell ad space to clients (i.e. the advertisers) knowing that the bulk of these ads will never appear in front of actual human beings - and making a killing in the process. Again, this involves faking impressions.

 

What are Impressions?

Here’s the deal - the online advertising industry is built on the concept of “impressions”. Now, popular belief suggests that an online “impression” is when an ad or a post is viewed by a person. However, in reality, an impression is defined as “a measurement of responses from an ad delivery system to an ad request from the user's browser”.

In simplespeak, an impression is registered as long as a browser serves an ad request. The ad may not load, it may be highly pixelated, it may be hidden from the computer screen, it may be hidden behind other ads, it may be stacked atop a bunch of other ads, or it may be infinitesimally small on the display. But if the ad’s request to display on a screen is accepted, it is counted as an impression. It has very little to do with humans or human eyeballs.

 

How Do You “Fake” an Ad, Though?

Besides the above-mentioned examples, there are many other ways publishers or ad agencies can dupe clients. Some infamous techniques are listed below.

  1. Pixel Stuffing: An ad is inserted into a 1x1 pixel - yes, just a pixel. So it’s invisible to the human eye, but it’s still technically “displayed” on screen, and is thus counted as an impression.
  2. Auto Impressions: When ads run within a webpage or app even when they are not being used.
  3. Ad Hijacking: Ad slots are hijacked by hackers to display different ads that the hackers get paid for displaying. Essentially, ad replacement.

 

How to Detect Ad Fraud?

It’s tricky. Simply because there are so many kids of ad fraud. But there are some ways to detect - and overcome - ad fraud:

  1. Spotting abnormally high click-through rates with minimal conversion
  2. Working only with trusted publishers with good reputations
  3. Paying for performance/conversion instead of paying for impressions/clicks
  4. Demanding updated and regular reports on campaign performance from ad companies + knowing the metrics being defined
  5. Hiring an expert to monitor campaign performance

 

Ad Fraud on Social Media

All major social media companies have announced measures against ad fraud - which is unsurprising since advertising is the bread and butter of this industry.

Facebook’s statements committing itself to act against ad fraud last week ring of déjà vu. It had previously taken action against app developers and marketers who allegedly generated fraudulent revenue via services running deceptive ads.

Google, for its part, has admitted that 56% of its digital ad impressions are never actually seen and informs advertisers on how to ensure ad visibility. Twitter routinely cleanses its platform of bots and publishes regular transparency reports. Instagram and LinkedIn have announced their own measures, primarily focused on bots.

But the fight between ad fraudsters and ad companies is likely to be a prolonged one. With every new protective update on social media platforms, fraudsters respond with more sophisticated tactics. Just as antivirus software companies regularly update their services to shelter users from newer strains of malware, so too will ad companies have to routinely improve their safeguards against ad frauds.

FIN.

There's so much happening that simply keeping up can be a task! How about a quick Quiz to help you retain? Subscribe to TRANSFIN. E-O-D and get them started.