In the emerging age of “responsible media”, you could be forgiven for thinking that marketers would want to exert more human control over production and placement.
But, increasingly, artificial intelligence algorithms are proving they can restore the primacy of ad creative.
That is what a host of industry executives discussed when they gathered on June 23 for the Global Forum on Responsible Media,
This video is a summary of interviews with executive who spoke in the creativity/technology advertising track presented by IBM Watson Advertising.
1. Dynamic creative rising
The New Majority: MediaCom’s Prabhu Aims To Make Advertising Addressable
Dynamic creative versioning is allowing advertisers to deliver a diverse range of re-mixed ad creatives for consumers. But Anush Prabhu – US Chief Strategy Officer and Global Chief Strategy Officer, Creative Transformation, for MediaCom – says companies need to lean on software for something that is becoming too complex for humans, in two areas:
- Production: Prabhu’s MediaCom is tapping tools like WPP Open and Flashtalking to produce creative in many versions connected to foundational insights.
- Optimization: Then he wants to understand which versions are working. “There are so many variations within those messages, whether it’s the right colour, do we have people in it?,” he asks. “How much of the product should be seen? All those aspects get even more complex when you add the different audience variations.”
2. Machines help scale creative palette
Robert Redmond thinks he has the answer – if producing a plethora of different ad creatives for a burgeoning range of audience types if complex for humans, call on the machines to help.
Specifically, machine learning like that offered by Redmon’s IBM is increasingly being called on to anticipate and remix the optimum ad creatives for different viewers.
“We teach an algorithm how to predict which individual assets to combine at real time to be most relevant for that consumer,” says Redmond, whose IBM Watson Advertising Accelerator assembles ad campaign creative elements based on audience reactions.
“We’re going to see more and more uses of technology and creativity together in very powerful ways to do this type of work.”
3. Context is back, with a fresh new look
‘Data Artistry’ Unlocks Context & Cohorts: Mindshare’s Clayton’s Post-Cookie Dreams
Creative-focused technology is important because there is a growing sentiment that ad creative, in the programmatic era, has been overlooked in favor of super-targeting alone.
But it also comes as ad buyers look for solutions in the era after third-party cookies and digital identifiers. And that is seeing the re-emergence of contextual targeting.
“Context has always been considered this old-school thing of the past,” says Sean Clayton, executive director, solutions officer at WPP’s Mindshare. “But, really, as you start understanding that people move in waves, they move in larger cohorts, the ability to start executing against those cohorts is actually pretty exciting, especially when you can look within the programmatic ecosystem.”
4. Restoring signal in an age of noise
Machine learning can help advertisers in the new world, despite declining usefulness of traditional identifiers, says Delphine Fabre-Hernoux, Chief Data & Analytics Officer at GropM’s Wavemaker.
“The power of machine learning is really to build this layer of intelligence on top of a more limited amount of signals and translate that into something which is quite meaningful,” she says.
“It may be insight, it can be intelligence that is going to optimise media planning, but it can also be the predictive piece. Everybody’s looking to really know where you need to put your media dollars to maximise the return on investment and contribute more to your bottom line.”
5. Piloting data signals
Xiao Lin of Xaxis wants to make sure clients have really bespoke creative that speaks to consumers. But he, too, wants to lean on technology to get there.
The GroupM division uses a tool called Copilot that uses signals like browser, location, time of day and the weather “to create thousands of creative variations on the fly”, Lin says: “It introduces thousands more different data inputs to which then our AI Copilot could actually optimise towards the output or the client’s outcome.”