Ad Verifiers Aim 'Surgical Strikes' Against Bots

Scott Knoll, Integral Ad ScienceIn the past week, Integral Ad Science and DoubleVerify - two companies whose aim is to ensure that brand advertisers get the kind of exposure and inventory they expect - unveiled their respective tools designed to limit, if not eliminate, the waste caused by fraudulent traffic – namely "bots."

The new tools, as outlined by Integral CEO Scott Knoll and DoubleVerify CEO/president Wayne Gattinella, each emphasize avoiding the usual solution of blocking entire site pages – something that only discourages buyers and sellers from taking greater action and can result in lost opportunities. Instead, they are each offering to strike only specific ad serving domains within a site while detecting the source of the phony impressions in order to shield ads before they are even served.

"The reality is, all solutions in the past have wiped out huge swaths of inventory," Knoll said. "That hasn't stopped the bots and it has only hurt publishers, while limiting advertisers' and agencies' options to reach the audience they want. So we just want to eliminate bad inventory with a scalpel, not a sledgehammer, which is what has in essence been the way it's handled. The way I see it, we just invented penicillin here."

The scourge of "nonhuman traffic" generating false impressions has been getting a great deal of attention the past few months. Despite the heightened awareness of bots by ad tech companies, publishers, agencies and advertisers, the problem shows few signs of going away. As such, companies such as Integral, DoubleVerify, AppNexus, Spider.io, Moat and Dstillery (formerly Media6degrees), have expanded their respective fraud protection and transparency wares that they say are designed to, at the very least, make it harder for fake impressions to worm their way through ad exchanges and networks.

In a sense, Integral's new "real-time fraud detection" reflects the shift toward predictive analysis from simply "verifying" after the fact if a brand's ad was served correctly when it rebranded from its previous name, AdSafe.

Integral's new bot detection is premised on preventing fraud via its firewall technology and is aimed at sellers and buyers. For its publisher clients, Integral seeks out patterns in vast volumes of historical browser and IP data. That allows it to screen a site's specific ad impressions in real-time, before it gets sent into auction, thereby stopping fraudulent inventory to buyers.

"There's no more need for buyers to have white lists, black lists or any other blocks that don't pinpoint the source of the problem," Knoll said. "If fraud is shown, the ad doesn't get served, so that gives buyers more confidence to spend their money. And a higher level of trust that leads to higher spending on publishers' sites is good for everyone."

So far, Adap.tv and Federated Media have been testing out the real-time detection. Walter Knapp, FM's COO, noted that they've worked on brand safety with Integral for years. In this instance, Knapp pointed to Integral's ability to score and rank individual URLs across hundreds of thousands of domains as the reason for taking an early look at the expanded services on bot blocking. "With a network as large and diverse as ours, it's important for us to have partners with the scale capabilities to handle our network," Knapp said.

DoubleVerify's "DV Bot Fraud Protection" tacks along similar lines. Neither entity views bot fraud as a distinct, business product line – at least not yet. Both companies include the new detection services as part of their larger set of services around impression quality and brand safety. DoubleVerify's Gattinella said that the company has no plans to charge for the service – at least until the end of the year.

At least for right now, bot fraud is a problem – but it's not the only problem facing the various sides of the online ad sales and buying spectrum, Gattinella said.

"Digital ad fraud overall is something that the industry needs to stay on top of because it's operated by very smart, crafty, fast players," Gattinella said. "The technology needs to constantly change to keep ahead of the fraudsters. Copyright infringement is also a significant issue. The goal is to maximize the quality of impressions that digital advertisers are buying in a way that helps the entire industry. Brand safety, viewability are also important and combating fraud is one part of that larger goal of making digital ad impressions more accountable."

There may not be much value in a singular expertise around brand safety. But there is value in the items that add up to providing assurances that advertisers are getting their money's worth on one end – and that publishers are worthy of the spending.

To put the issue of bots and security in further context, Knoll pointed to Integral findings that say that 20 percent of display ad impressions purchased through exchanges are "suspicious of being fraudulent."

Video advertising, which is becoming more important to brand advertisers, is experiencing a particular rise in attacks from bots – due to the rising dollars that are flowing into the space via exchanges, which is why companies like Adap.tv are interested in being seen as offering safety. In addition to Adap.tv, video ad tech company YuMe, which is primarily focused on brand advertisers, also released its own fraud detector within its Placement Quality Index. The discussion among video ad tech providers started in earnest this summer, when Tremor Video CEO Bill Day issued a general message to the industry professing his company's commitment to transparency. Aside from zeroing in on fake impressions and click fraud, YuMe's special sauce to combat bots also comes with "ad stacking," where ads are tucked into a single out-of-view or below the fold ad.

While only three percent of pre-roll ad impressions are tied to bots, the majority of fraudulent activity on video ads is seen across in-banner video ads, where up to 40 percent of impressions are tainted, Integral's Knoll noted when asked about the sneak attacks on video impressions.

"We've been working with Adap.tv, as well as FM, on the general issue of brand safety for a while," Knoll said. "But we realized that there's a lot of value in the reach that we have. And right now, there's a lot of attention on bot fraud. There are a handful of companies that can do this in real-time. We have advantages because we see a couple billion ad impressions a day, which a lot of companies don't have."

2 Comments

  1. Shawn Lundt

    These guys are relying on IP for bot defense? That's like arming infantry soldiers with high-powered squirt guns. The methodologies employed by downstream validation firms were never intended for security and they are mocking the industry by suggesting they are credible

    Reply
    • I don't think they mean IP for bot defense, but rather the data you get off each time a certain IP Address comes into question. I believe a lot more goes into this then just blacklisting an IP Address.

      Reply

Add a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>