Fighting ad fraud: 5 questions all advertisers need to ask

For digital advertisers, the goal is simple enough -- you're after customer engagement, and maybe most of all, scale.

But, as online ad ecosystems have evolved, so too have opportunities for fraud, threatening the value of both. A wide variety of sources indicate that anywhere from 15 percent to 61 percent of all display ad traffic on a global basis is driven and maintained by robots -- not humans. These bots imitate real human behavior, delivering phony scale and engagement to scam brands out of billions in ad dollars.

If, for a moment, we can forget the large range, we can at least agree on two things: 1) it's great business for the folks that operate botnets, and 2) it's a huge problem for advertisers. Not only is a huge waste of ad dollars and spend, but it's a huge analysis issue -- preventing advertisers from fully understanding a campaign's ROI, value, and effectiveness.

And as automated buying and selling via programmatic gains greater adoption, fraud is only getting worse.

So what are advertisers to do? How can they navigate fraud? Here are five questions any advertiser or brand needs to ask when considering potential partners for a relationship.

How do you access inventory?

Today, more ad networks, exchanges, and SSPs are getting their inventory secondhand. They're sourcing from other networks, exchanges, and SSPs rather than building direct relationships with publishers themselves. Why is this problematic? The more players there are, the less control and insight the original partner will have over inventory and placement, and the greater the chances that a bad actor appears somewhere along the line. Having direct relationships is always preferred, ensuring -- for the most part -- that partners know what they're getting.

What monitoring tools or practices do you have in place?

Every partner you consider should be challenged to outline the monitoring tools and practices they have in place to ensure that they stay clear of fraud. Your partner should have a combination, if not all, of these to safeguard your brand and your dollars:

  • Domain-level reporting that enables the partner to approve every publisher domain before ads are ever served
  • Contextualization engines which seek out unacceptable keywords and risqué topics to prevent ad placement next to undesirable content that is not aligned with the brand
  • Tracking platforms which monitor unwanted activity like fraudulent clicks, traffic purchasing, adware, spoofing, etc.
  • Ad exclusion properties intended to prevent serving to offending IPs or user agent strings identified by the IAB's known list of spiders and bots or other tracking methodologies

What criteria does your publisher partner adhere to?

It's important to understand what type of inventory you'll have access to by working with a partner. While you can't check every site, understanding the partner's publisher acceptance criteria is really the key to clarity. The best partners will have rigorous conditions in place to deliver quality, covering counting metrics like monthly unique users (the more, the better), in combination with more qualitative elements like brand appeal, user design, and ad placement.

Do your publishers distribute illegal content?

This may sound like common sense, but it's an overlooked guideline, given the prevalence of pirated content online. Some networks and exchanges actually work with sites that enable access to illegal content under the defense that the sites themselves do not host the illegal material. This is, as I see it, a reckless distinction, and a good partner will refuse to enable this type of activity by monetizing those contributing to these practices. At its core, it's really a brand safety issue that needs to be addressed early on in conversations. Safe and transparent partners will not have issues providing you with comprehensive site lists, which should help mitigate serving on illegal content.

How accountable are you?

The ad ecosystem can be messy, making it easy at times for partners to write off issues when they do occur, attributing it to an inherent institutional problem. But advertisers don't want to hear that. If a partner isn't fully accountable, the problem doesn't get fixed. From the outset of a relationship with any partner, advertisers should have candid conversations on accountability and how to respond to problems if they do occur to avoid longer-term issues. Advertisers do need to understand, though, that we need to work together in order to adequately shift the paradigm back and limit the rise of non-human traffic.

For advertisers and brands seeking a partner and hoping to avoid fraud's pitfalls, these are five great questions to ask. They won't help you avoid fraud completely, but they serve as a good starting point when assessing how your partner can collaborate with you to take on this growing issue. Others in the space -- what else would you recommend?

Craig Simmons is the manager of product strategy and operations at Exponential.

On Twitter? Follow iMedia Connection at @iMediaTweet.

 

Comments

Armando Carrillo Jr
Armando Carrillo Jr June 17, 2014 at 1:21 PM

Great article. The "What monitoring tools or practices do you have in place?" section is on point with many of the challenges zVelo's ad-tech partners face with regards to domain-level reporting and ad inventory vetting. zVelo's contextual categorization engine helps vet ad inventory to make sure web pages are brand-safe and to ensure ads don't appear next to objectionable content. The engine is also licensed to help make sense/visualize big data. Another area we're tackling is the detection of illegal/pirated content. That in itself presents a HUGE challenge.