Ad fraud: a guide to detection and combat
Published: June 2, 2017
Author: Luis Valles
Ad fraud is a costly issue for online advertisers today. Some estimates predict businesses could spend up to $16.4 billion in 2017 on fraudulent traffic. This is a serious problem, and one that everyone should take note of. What can you do to avoid falling victim to it? Let’s dig in.
Types of Ad Fraud
First, let’s look at the types of ad fraud advertisers may encounter:
Ad stacking: This is seen when a publisher has multiple ads stacked in one placement on a page, causing several impressions to serve while only the top ad is seen. This tactic can also count clicks for all ads in the stack and can skew viewability numbers.
Pixel stuffing: In this case, a publisher places ads into a single pixel on the page so users cannot see it but impressions are counted.
Traffic fraud: Pretty straightforward: publishers purchase bot traffic to serve more impressions on their sites and boost their volume numbers.
Conversion fraud: This happens when companies hire people or bots to interact with ads to drive leads that will never convert. Bots are becoming smarter to the point where they can complete more complex actions beyond just clicks. An example of this would be a “user” filling out a lead form for a student loan but never interacting with the site again.
Ad injection: This is where fraudulent ads appear on sites through browser extensions, unbeknownst to the viewer or publisher, and cover the existing paid ads so that they aren’t visible.
How to Identify Ad Fraud
Now that you know the types, here’s a breakdown of how you can identify fraudulent activity. (It’s important to detect it early on, because ad fraud can be extremely costly.)
The first thing we look for is unusual click or conversion activity. This can come in the form of:
- Influx of clicks or extremely high click-through rates
- Increased number of clicks while conversion remains flat
- Repetitive clicks from same IP addresses
Next, we look for ad engagement or traffic coming from unusual places such as:
- Heavy traffic coming from third-world countries
- Traffic coming from outdated web browsers
- Activity coming from areas outside of your target area
Another indicator of ad fraud would be high rates of shopping cart abandonment:
- Bots and click farms can complete lead forms but cannot purchase yet
Finally, a red flag would be bid win rates over 100%, if you are running on a self-managed platform like The Trade Desk or DBM. You can’t win more impressions than you bid on, so if you are seeing this, something fishy is going on.
How to Combat Ad Fraud
The good news is, as fraudulent publishers and bots get smarter, so do we. Here are some best practices for combating ad fraud:
- First, you’ll want to constantly monitor performance data to catch any suspicious activity. If campaign performance is looking too good to be true, you may want to dig a little deeper into where the activity is coming from.
- Next, update your blacklists to ensure ads are not being served on suspicious sites. This is more of a reactive way to stop fraud, but it can help stop your ads from appearing on fraudulent sites moving forward.
- If you are not already geotargeting, block countries that are known to have high fraudulent activity.
- Another best practice is to work with sites, networks, and exchanges that are transparent and use fraud monitoring tools.
- You can also consider on-boarding anti-fraud tools like Integral Ad Science or White Ops that are MRC (Media Rating Council) accredited to block fraudulent activity at the pre-bid level.
Finally, you can work with an agency like 3Q Digital. We know the warning signs and work tirelessly to ensure our clients do not fall victim to malicious activity. Contact us if you have any questions about ad fraud or would like to partner with us on your paid advertising needs.