Marissa McArdle, Nielsen’s global vice-president of digital audience measurement, said brands need to consider “what the actual, viewable demographic is versus applying an on-target percentage to viewable impressions”.
To beat the bots at their game of adaptation, utilising artificial intelligence and machine learning in programmatic buys can come in useful in the fight against invalid traffic.
“Make sure that you’ve got some robust [machine] learning – fraud detection – running on your advertisements,” McArdle said at the recent Asia Pacific Media Forum in Bali, Indonesia. (For more about how to beat the bots, read WARC’s report: Ad fraud is eating up your KPIs – what to do about viewability issues.)
That said, machine learning can only complement ad fraud detection, as there is no substitute for real human intervention. Staying vigilant is important as trends won’t always appear across data sources.
“You cannot expect to be able to understand the true ROI of something if your underlying data is corrupted,” McArdle said.
Measurements can allow marketers to react quickly by tweaking their placements accordingly. “Make sure that you’re taking a look at your metrics daily, optimising, changing your buys, adapting to what you’re seeing, and moving the data from placements that are less effective to placements that are more effective,” she said.
To move forward as an industry, marketers and publishers have to work together as an ecosystem without losing sight of the collective end goal, she added. “We need to focus on audience. We need to work together. We are better together.”
Sourced from WARC