When IVT Is Quite Valid

Just the mention of the term invalid traffic (IVT) elicits hisses across the digital advertising ecosystem. As the programmatic advertising ecosystem evolved—and buckets of advertiser spend were dumped into DSPs—a barrage of bots ensured countless ads went unseen by human eyes. Scores of premium publishers were cheated out of impressions (and revenue!) and advertisers were defrauded of their investments.

A lot has happened since then—most important, the development of third-party IVT measurement solutions—but the animosity and confusion lingers on. The terms IVT and ad fraud are often used interchangeably. The only good bot is a dead bot!

Well, that’s kind of a silly statement because bots aren’t alive to begin with… And also because IVT is not just about scammy bots. There’s a great deal of benign IVT roaming the web—much of making the Internet better and safer for the end user. How do you think Google became the dominant search player it is? Superior crawlers—a classic example of General Invalid Traffic (GIVT). 

Another great example? The Media Trust’s scanning—we use emulated users to determine what third-party code is running on websites as well as seek out malware and malvertising schemes as they evolve in the wild. Those expeditions help us build robust block lists and guard publisher properties.

IVT’s dual nature

You see, IVT comes in two flavors, according to the Media Ratings Council (MRC), which finalized a significant addendum to its IVT measurement standards in October 2020.

General Invalid Traffic (GIVT) is the benign and often beneficial stuff! Useful tools include:

• Malware scanners
• Crawlers
• Bots used to detect unsafe browsers
• Self-identifying bots
• Routine maintenance bots 

Sophisticated Invalid Traffic (SIVT) is bad! Examples include: 

• Automated browsing from a dedicated or non-dedicated (typically hijacked or infected) device 
• Incentivized human invalid activity (e.g., paying real people to click on ads) 
• Manipulated activities like mobile redirects or tricking users to click (e.g., fake close window buttons) to open new windows
• Falsified measurement events like visit, impression, viewability, click, location, consent strings, etc.

See? Not just bots—there’s plenty of human-driven invalid traffic to go around. 

Recently, buy- and sell-side measurement provider Integral Ad Science (IAS) dropped the term “Fraud” from its well-read reports and embraced the more inclusive Invalid Traffic label instead. The third-party measurement provider stated that it aimed to “create a clear distinction that invalid traffic is not always malicious—some of the time, it’s benign non-human traffic (GIVT).” 

In accordance with MRC IVT measurement standards—which need to be followed to receive accreditation—certified IVT measurement services must subtract GIVT from the overall IVT count for reporting and billing in the ad ecosystem.

However, a lot of bundled IVT numbers—that is, GIVT + SIVT—are tossed around the industry, garnering super-sensational headlines. “It’s all bots, BOTS! No ads are ever seen!” No surprise that many in the industry—particularly media buyers—still associate all IVT with “bad bots” and “fraud.” 

Getting recognized

Thing is, a lot of GIVT—such as well-known crawlers—are easily identified with domains shared on public lists. They’re easy to filter out and avoid serving ads to. 

The Media Trust doesn’t have that luxury—if we made our tracking domains and IP addresses publicly available, bad actors would jump all over them and make our protective measures absolutely useless. Even if we did expose ourselves to these lists, platforms would opt to not serve ads to our bots, which would mean we would stop seeing programmatic ads and our effectiveness would plummet. 

We tirelessly work to obscure details that identify our Media Scanner to bad actors so that we can offer the most accurate view of malware across the ecosystem, and find the sophisticated threats that actively evade known blockers and scanners (which of course fuels our Media Filter real-time blocker). 

So we work directly with publishers and major ad tech players to help them identify The Media Trust’s traffic—on a post-impression basis as well as in real time. This includes several MRC-certified measurement providers. The last thing we want is to aggravate our publisher or ad tech partners’ headaches around IVT. 

But say a partner comes your way with some IVT numbers that seem larger than life—here are a few simple steps to take.

• Ask what measurement company is the source of the data, and whether they delineate between GIVT and SIVT. Note that MRC-certified vendors are supposed to be able to identify and then remove GIVT from their IVT reports.
• If you can or are willing, show your own traffic measurement with and without GIVT stripped out. Chances are, the large number will correspond with your traffic that includes GIVT.

Originally published on The Media Trust's blog.