Blog

March 30, 2016

4 things brands should consider when going Programmatic this year

By: Martin Bromfield

Smart creative executions have more impact when delivered to the right people, under the right conditions. At comScore, we break this down into four areas that combine for what we’d call the “verified impression”. More simply, it’s the question advertisers want to know – could my ad have been seen by my target human audience in a brand-safe environment?

The four pillars are Invalid Traffic (IVT) detection, viewability, audience reach and brand safety. Whilst they are distinct and detailed topics individually, having all four under control leads to an end result greater than the sum of its parts.

We believe these factors remain fundamental, regardless of how advertising is bought, and that advertisers should take the time to understand how they can control them in a programmatic world.

Invalid Traffic Detection: “Is my ad going to a human?”

IVT is a primary concern that wastes money at source (ads delivered to non-humans) and skews ensuing optimisation and performance measurements. The IAB estimates 36% of traffic is bogus, and while not all IVT is deliberately fraudulent, this large and growing component involves a diverse spectrum of bots, click farms, browser hijacks, ad stacking, content scrapers, domain laundering and more.

Although the vehicles are different and evolving, they share the goal of channelling legitimate advertising revenue into fraudsters’ pockets in an undetected manner. The use of hijacked machines posing as humans is particularly worrying for device owners (who have no knowledge it is occurring), and because these bots are so difficult to recognise, a huge concern for advertisers.

Sophisticated fraud can only be detected and removed by equally advanced, dedicated tools with an understanding of real human behaviour.

Viewability: “Did my ad have a chance to be seen?

As logic dictates, ads that had a chance to be seen have more opportunity to make an impact – the MRC in the US defines ‘viewable’ as 50% of the ad’s pixels being visible for over 1 second.

A key challenge here is managing discrepancies between what different tools measure. For example, is invalid traffic removed before viewability is calculated? Is the tool measuring all open browser windows or just the one on top? Can it spot ads served outside of the viewable window or multiple ads that are stacked on top of each other?

Let’s take the example of Invalid Traffic. Not removing it before calculating viewability will artificially inflate reported viewability (by including ads ‘seen’ by bots). This in turn reduces ROI, as more impressions share credit for the overall response, undervaluing performance of the legitimate ones. It is also impossible to reliably compare this data against results that did remove IVT.

Clearly it’s important to understand what is being measured, and ideally that’s stringently-reported viewability that can be combined with the removal of IVT.

Audience Reach: “Did my ad go to the people I wanted to reach?”

Targeting is a big expectation of programmatic advertising. For similar reasons to viewability, it’s important to remove IVT before judging effectiveness, particularly fraud based on realistic human profiles which have zero chance of becoming customers.

Evaluating targeting ideally involves a secondary source that does not rely on the same cookies or data used to deliver it. On a more positive note, so long as they avoid IVT, impressions which fall out of target still have an opportunity to be effective.

Brand Safety: “Was my ad surrounded by suitable content?”

Environments that are ‘safe’ vary by brand. The range extends from usual suspects such as pornography, violence and racial hatred to more nuanced issues such as cruise ship ads appearing in news stories about cruise ship accidents.

As well helping to avoid legitimate errors, issues such as domain hijacking or the masking of URLs (where it appears that an ad is going to a safe environment but in reality, it’s not) require sophisticated fraud detection to avoid, especially during rapid inventory trading on programmatic platforms.

“So, what should I do?”

Addressing all factors in isolation is a step in the right direction, but the real benefit occurs when you combine them into a single source to deliver “verified impressions” - an unduplicated and validated measure of clean impressions that had the opportunity to make an impact with your audience. This can be fed into bidding strategies and optimisations that will continue to make your digital advertising more valuable, and keep your marketing budgets out of the hands of criminals.

Far from being a negative chore, this is a positive process that makes clear which impressions are doing the work. It’s likely they are working a lot more effectively than you might have thought.

Tags: Ad Impressions, Ad Validation, Advertising, Programmatic Buying, Viewability