'Programmatic is too simplistic'

Rosie Baker
By Rosie Baker | 19 April 2016

This article first appeared in AdNews in-print. Click here to subscribe to the AdNews magazine or read the iPad edition here.

Programmatic is too simplistic and the technology is undermined by the measures it uses to inform targeting. That’s the bold statement from Dr Karen Nelson-Field, CEO at Media Intelligence Co., who says that machine learning can tackle the simplicity inherent in those measures.

“The measures that underpin programmatic are so simplistic,” she said, adding that while the technology might be sophisticated and getting even more so by the day, part of what it can achieve is undermined by relying on demographic data. She advocates moving away from persuasion measures and explicit memory to real-time and implicit measures.

Nelson-Field is an academic with a heritage in digital. She has made a career out of challenging the norms and questioning what is often accepted as truth. She’ll be our keynote speaker at the Media Summit on 19 May.

You can buy tickets for the Media Summit here.

“There’s this whole concept of recency which says the closer to the purchase occasion, the more impact the ad exposure has … What programmatic does is hit people closer to the purchase occasion, but no one has nutted recency in programmatic – but it’s coming,” she warned.

“My point is that a lot of these measures are still relying on pre– and post–attitudinal changes. It’s from 1965 so I think it’s like taking two steps forward, one step back. If I was buying media I would still use programmatic with demographics; it’s efficient, but the concern is that a lot of marketers over-target, thinking that wastage is a bad thing.

“What super-targeting does is it removes the opportunity for growth in the future because you’re never getting new people to consider your brand.”

Nelson-Field, who is about to launch a joint venture with machine learning company Jemsoft, is building an AI platform that takes into account more measures than simple demographic data to learn how to predict brand growth.

“At the core – it’s computer vision that learns,” she explained. “I’m applying that to brand scanning and integrating it into pre-testing. It’s looking for variables that are aligned to what we know about how brands grow.”

You can buy tickets online here 

Have something to say on this? Share your views in the comments section below. Or if you have a news story or tip-off, drop me a line at rosiebaker@yaffa.com.au

Sign up to the AdNews newsletter, like us on Facebook or follow us on Twitter for breaking stories and campaigns throughout the day. Need a job? Visit adnewsjobs.com.au.

comments powered by Disqus