online-ad-fraud-detection

Ad fraud is a billion-dollar industry. Let me say that again. Ad fraud is a BILLION-DOLLAR INDUSTRY. Unfortunately, as programmatic ad buying becomes more prominent in digital advertising, ad fraud will grow with it.

Here are some stats from a 2014 study by the Association of National Advertisers and WhiteOps that should ring an alarm:

  1. Up to 50 percent of publisher traffic is bot activity.
  2. $6.3 billion of the digital advertising industry will be based on fraudulent activity in 2015.
  3. Between 3 percent and 31 percent of programmatically bought ad impressions were found to be from bots, with an average of 17 percent.

Source: Adweek

If we’re not careful, digital advertising could just be a great waste of money, and mobile advertising, which has been largely ignored by fraudsters in the past, will be overrun with malicious activity. Even though 100% fraud prevention is nearly impossible, fraud detection and transparency should be instituted on every advertising platform that exists.

Let’s dive into different fraud techniques and some ways the ad tech industry can tackle ad fraud.

How Fraudsters Are Attacking Advertising

Sadly, there are a myriad of ways to trick advertisers into thinking real people are actually engaging with their ads. Over the years, these techniques have evolved as consumer technology and ad tech have evolved. Just like programmatic technology gives advertisers the ability to scale, computer bots give fraudsters the ability to scale their activity. Here are the main techniques:

Click Fraud/Bot Behavior – The most popular fraud technique is click fraud, which is simply a person or most likely a bot continuously clicking an ad without having interest in the ad. This leads into sophisticated bot behaviors, which include watching videos, playing gaming apps, and even filling out lead-gen forms! Bots can replicate human behavior, which makes digital advertising daunting.

Mobile Location Data Spoofing – Mobile location data is crucial to mobile advertising, and inventory with this sort of data costs more. Fraudsters use fraudulent locations to waste more of advertisers’ money. When location data is questionable or fraudulent, mobile advertising campaigns don’t fare too well.

Domain Laundering – With domain laundering, advertisers create fake sites or use an existing site that’s unappealing to advertisers and provide the inventory to ad exchanges while masking the identity of the site. This tricks advertisers and ad exchanges into thinking the inventory is legitimate. Read more about domain laundering.

Ad Stacking – Ad stacking is placing multiple ads on top of each other in a single slot, with only the top ad being viewable. This is a form of impression fraud because the advertiser is paying for impressions even if the end user is not seeing an ad.

Redirects – This doesn’t affect the advertiser as much as it affects the end user. I included this because it’s one of the shadiest tactics plaguing the digital advertising industry and it needs to stop. Some “advertisers” use this to take the end user to a site they aren’t expecting (sometimes polluted with pop-ups). This can lead to spam and viruses.

Give Advertisers More Transparency And Detection Tools

We need to prevent fraud, not just detect it. Everyone’s mind should be shifting to prevention. However, no ad inventory is going to be 100% fraud-free, just like crime will never be completely eradicated. If we shift our focus to fraud prevention instead of just merely detection, we can reduce ad fraud significantly with really robust detection tools that’ll facilitate prevention measures.

Now that we know some of the ways fraudsters steal advertisers’ money, let’s discuss some ways we can help advertisers fight back.

Data Sharing With Advertisers – The best way to make advertisers feel better about the safety of their advertising dollars is to allow them to investigate and verify how much of their ad traffic is fraudulent, and then helping them take prevention measures. One way to do this is allow them to export all the click and conversion data from their campaigns in a .csv file.

On Decisive, we call this Click Explorer, and it gives advertisers the ability to analyze their clicks to detect possible click fraud. It’s important because it allows us to identify if there are specific patterns of bot clicks from specific IP addresses, user agents, or placements, which leads into the second phase.

Decisive Ad Fraud Protection

Global Blacklisting – Once the advertisers know the IP address, placement, user agent, etc. of the suspicious activity, the attribute is blocked platform-wide to ensure no other advertiser experiences the same activity.

Create Benchmarks – Use platform data (customer blacklisting, performance data, click data, etc.) to create benchmarks.

Partner Feedback – Advertising platforms should work with inventory partners on flagging suspicious inventory and providing information for future prevention.

Kochava Fraud Protection

Screenshot of Kochava’s fraud protection site.

Detect Trends – If fraudsters are going to be sophisticated, you have to be sophisticated as well. This means using a little behavioral science to detect fraudulent patterns. One example of this is analyzing “Mean Time To Install (MTTI)”, a term created by Kochava, a mobile analytics company. “Mean Time To Install” is the average time it takes between when a user clicks to download an app and when that user launches it for the first time.

Detection science such as Kochava’s can be game-changing in the fight against ad fraud because it goes beyond just knowing what exists; it detects how fraud manifests.

More Industry Oversight

Lastly, there needs to be industry oversight and accountability.

The IAB Trustworthy Accountability Group is an industry accountability program designed to fight ad fraud. In their own words, “the organization will operationalize principles around the critical issues that face the industry in this arena” and “monitor the ecosystem for compliance and develop incentives for broad industry participation and consequences for untrustworthy actors.” The war against ad fraud won’t be won unless every player in the industry takes a stand, especially the major players.

All in all, the fight against ad fraud will be long and tedious and may never end. But with a little bit of effort and a shift in mindset, we can make sure that advertisers’ money stay as safe as possible.