Integral Ad Science is one of the third parties Google is bringing in to help monitor ad placements on YouTube in the wake of advertiser backlash over brand safety concerns.
The 400-person firm topped the $100 million revenue mark last year and works with more than 4,000 publishers, advertisers and ad tech vendors, including Facebook, AOL, Nielsen and TradeDesk, in addition to Google.
Ahead of the release of the company’s Media Quality Report for the second half of 2016, Marketing Land spoke with Maria Pousa, Integral Ad Science’s CMO, last week about recent events, including the Google ad boycott, whether the industry is at a turning point and how the company anticipates advertisers will start thinking about evaluating their digital ad buys.
On YouTube and brand safety
IAS and the other third-party monitors like Moat, Double Verify and comScore are emerging as winners in the wake of advertiser concerns over fake news, ad fraud and walled gardens, and the recent protests over ads being shown alongside racist and extremist content on YouTube.
Is this a real turning of the tide in which the YouTube crisis is just the last straw? Will there be a real course correction in the way the industry approaches programmatic and brand safety, as the IAB’s Randall Rothenberg since declared in a letter Tuesday? Pousa says it’s still too early to tell.
She knows there has been a cycle of flare-ups around brand safety that the industry has been dealing with for years. “What’s changed,” says Pousa, “is the environment. Usually, the attention was from within the trade, now it’s moved to mass media outlets. First it was fake news, now it’s The Times [coverage of YouTube]. We’ve been getting calls from mainstream press — from politics, society and business press — that don’t usually cover these things.”
With the growth of programmatic, some media buyers and brands have focused on audience and reach more than on contextual experience in order to maximize media spend. “Some might be reconsidering the right place for the brand,” says Pousa. “The level of confidence is different from brand to brand. Some look for scale and reach and feel like they can weather the storm.”
But, she says, “I am not sure this is going to go away this time. It’s tied to a socio-political climate with global news impact. Even our Australia office, our Singapore office, is getting calls about [YouTube and brand safety].”
Pousa has commended Google for taking steps to address the issue and look for a solution. Whether this is truly a turning point for the industry, Pousa says, let’s chat again in six to eight weeks.
IAS’s Media Quality report analyzed brand risk across programmatic campaigns vs. publisher direct campaigns that run through its systems. (IAS looks at seven content categories: adult, alcohol, hate speech, illegal downloads, illegal drugs, offensive language and violence.)
While noting the current climate, the report also found that the volume of impressions flagged on objectionable content in the second half of 2016 in the US was actually down from the first half of the year from 9.5 percent to 8.6 percent across all buy types. In the second half of 2016, 9.5 percent of impressions were flagged for brand risk content on programmatic campaigns. For publisher direct campaigns it was less than half, at 4.6 percent. Adult content, offensive language and violence were the top sources of objectionable content flagged for both buy types.
Source: Integral Ad Science, H2 2016 Media Quality Report
On Methbot and ad fraud
In January, White Ops reported a giant botnet that spoofed premium publisher domains on which to capture video ad revenue. It was estimated to be siphoning $3 million to $5 million a day from publishers. That impact was disputed by many in the industry who said they’d quickly identified and isolated the threat. Pousa says it was a good PR effort, but “From our perspective, we see billions of impressions and only saw 0.03 percent of impression volume [from those URLs]. And others had a similar response. It was basically insignificant.”
It had the effect of scaring customers without real cause, says Pousa. “We see bots all the time and are constantly fighting an arms race. We see the same types of behavior from other bots.”
Still, the Methbot news can be viewed on a continuum of heightened awareness about fraud, fake news, Russian hacking and hyper-partisan propaganda on sites like Breitbart since election season. Methbot wasn’t just perpetrated by an old hacker group; it was perpetrated by Russian hackers. That fact likely added to the traction that report gained.
The IAS report does show several trends in ad fraud. One is that fraud is 88 percent higher on home pages than on article pages, but that the likelihood of fraud increases as articles age and bots have time to discover them. Two is that fraud risk increases between midnight and 4 a.m.
On optimizing for time and frequency
“If we start looking at time as the currency, we can start doing platform and publisher standards of time exposure. That starts changing the conversation,” says Pousa, because the MRC viewability baseline of one second is clearly not a indication of effectiveness.
The IAS report finds that most campaigns leave the vast majority of consumers underexposed to messaging, with 60 to 80 percent of users getting served just one or two impressions and 70 to 85 percent served just one viewable impression. Further, 50 to 65 percent of users are exposed to ads for a total of five seconds or less.
A new focus on time exposure, coupled with heightened awareness of brand safety, could reframe the way advertisers buy and evaluate digital media. IAS expects “advertisers will move past impression- and media-based metrics to consumer-based measurement,” with time exposure being the “new digital currency.”
Source: Integral Ad Science, H2 2016 Quality Report