The study also cites a 400 percent rise in so-called “aberrant traffic” in actions like registration and voting, resulting in a costly nuisance for both publishers and advertisers. Solve Media says it conducted the 18-month study on the basis of 2 billion authentications performed across more than 5,000 publishing sites.

The upshot, according to Solve Media, is that advertisers may be directing up to 10 percent of their online advertising dollars at this fake, robot-driven traffic. Meanwhile, publishers are burdened with purging bots from their comment boards and registration systems.

Solve Media seeks to solve this problem by helping publishers keep the bots away with ”Captchas” — those pictures or squiggly letters that require you to type something.

According to Andy Wilson, SVP of Digital Marketing at Meredith Corp, the Captcha tools have reduced the number of bots entering contests. In the past, he says, bots entered to get their babies on the cover of Parents magazine or to win $25,000 from Better Homse & Gardens.

Ari Jacoby, CEO of Solve Media, says that it’s important for publishers to aggressively battle the bots.

“Once bots begin to attack a sight, they’re everywhere on the site … But once they meet significant resistance, they’re trained to go away.”

Solve Media also claims that its Captchas are calibrated to be easy for humans and difficult for bots. This means that a real person is less likely to encounter an indecipherable squiggly line but instead will get a straight-forward message to type like “Toyota Corolla” (which presents, of course, another advertising opportunity).

Finally, here’s an infographic that Solve Media is using to present its study findings:

(Image by   Principal via Shutterstock)

The findings are part on a white paper that summarizes the ad experience of twelve major brands including Sprint (NYSE: S), Ford and Kelloggs.

In a preview of the paper released in January, comScore reported that 31 percent of the 1.8 billion ad impressions purchased by the companies were not seen at all.

The findings are significant because brands are shifting a greater percentage of their marketing budgets to digital advertising. And, unlike a TV or radio ad, it’s hard to tell when and where an online display appeared — or even if it appeared at all.

Today’s white paper shows that brands are struggling not only to make sure their appear, but that they appear to the right viewers.

According to the study, 3-10 percent of all ad impressions are shown to non-humans — bots and spiders that induce a web page to display an ad. It also finds that 72 percent of all ad campaigns resulted in brands placing their ads next to “objectionable” content. In practice, this means that brands paid to have their ads appear on sites dedicated to file-sharing or phishing:

These findings may worry ad buyers but may actually symbolize overall progress for the still-emerging online ad industry. That’s because this is the first time, according to comScore, that brands have been able to validate their ads across a variety of ad services.

The findings may also presage a shift in how brands and publishers approach ad buying. According to Anne Hunter, an SVP at comScore and one of the study’s authors, ad-buying will become less a leap-of-faith for advertisers. Publishers, meanwhile, will have a greater ability to ensure they don’t off-load premium ad inventory at discount prices.

The study also found that, for now, “good guys” are not being rewarded for providing a quality advertising audience — brands have been paying the same amount for an ad regardless of whether it is seen or not.

comScore has an interest in the findings because it sells ad monitoring tools. The company gathers its research by working with a large sample group of internet users who let comScore track their web surfing behavior.

comScore’s report can be found here.