Nothing promotes more conversation, speculation, excitement and fear right now than the subject of AI.
And so a cross-industry audience packed into the auditorium at London’s Soho Hotel to hear AI practitioners and experts reveal what they’ve learned so far, how to get started with AI — and what they predict.
For Nick Baughan, director of global agencies at Meta, AI could potentially power an outcomes-based renaissance in the platform partner economy, he told the audience in an opening provocation.
“This has determined Meta’s investment roadmap for the last 10 years and we have invested meaningful CapEx into AI to be leaders in this space. This long-term investment is now delivering our Advertisers 32% more return on ad spend, and 17% improvement in cost per conversion.”
But Baughan stressed that the point of the Global Pioneers program, in which Meta and Campaign have been looking at advertising’s future over the past three years, is to share experiences and to learn. He couldn’t say with any certainty what would happen with AI even in the next six months, he admitted.
Here are the key take-aways from Baughan’s discussion with Amir Malik, managing director of Accenture, Alex Dalman, managing partner/head of social & innovation at VCCP & Faith and Niel Bornman, chief executive officer at Publicis Media UK.
From hours to outcomes
Baughan offered a challenge to the old adage in the marketing and advertising industry that you can have something good, you can have something fast, and you can have something inexpensive; but you can only pick two.
A shift away from commissions based on hours towards an incentive model based on positive outcomes could remove many of the constructs that limit advertisers from achieving all three of these desires.
AI and its potential applications could make this entirely possible, Baughan hypothesized.
Who pays?
Alex Dalman, managing partner/head of social & innovation at VCCP & Faith (VCCP’s AI creative agency), brought the conversation back to the present by issuing some observations on the practicalities of using AI.
Whatever the future promises, the present is more gnarly, she said. Faith was set up eight months ago to get on with doing and actually understand these tools.
“A lot of people assume it’s definitely going to be cheaper and easier. But there is a cost to generate on the platforms and someone will get a bill someday,” she said. “It’s not free. It’s not magic.”
Data is another area of potential problem, namely in the kind of training data necessary for AI’s large language models. The agency works with O2, but even its enormous cache of copy could not train the AI on tone of voice and Faith had to go back to go forward, writing new copy to train the model.
Data dreams
Niel Bornman, chief executive officer at Publicis Media UK, went further and said that selecting the wrong data can create bias and “hallucinations in the AI.”
He believes companies are at “an inflection point” where, if they haven’t organized their data, they will struggle. Even large companies only have access to their own data, so collaboration becomes crucial.
Brands should also think about data as it relates to brand character, such as photography style, or voice over artists, Dalman said.
“All the data that has created the brand world: you’ve now got to somehow grapple that back and use it to train models,” she said.
She predicts AI buyouts of things such as voice overs, illustration and photography, will be a big issue going forward. “You have to define your brand really distinctively otherwise the AI will just spit out something pretty generic.”
The outcomes will be entirely determined by the skill and data used to train models and prompt them, she warned: “crap in, crap out.”
What can AI do for you?
Amir Malik, managing director at Accenture said that while marketers anchor the conversation in a media-centric way, AI is disrupting the entire business.
“As an organization, you need to understand where you are on your journey,” he said. “Most companies will achieve differentiation by getting AI to the edges of the organization and helping everyone make better decisions.”
At VCCP, a global training program encouraged people to lean into the technology and removed some of the fear, said Dalman.
“We ended up with so many use cases, particularly from the creative side where [AI] helped visualize ideas quicker and helped sell in some of our more ‘out there’ advertising. The flying goat for Virgin Media started as a mid journey scamp,” she revealed.
Rather than fearing AI as a job destroyer, organizations should see it as a way to become better and more productive, said Mailk. “We have a massive shortage of skills in our industry. I don’t see AI replacing people but allowing us to keep up with the demands.”
Where next?
Bornman said the marketing industry has a three-to-five year window to bring together all of the component parts of the marketing system and to understand how AI will impact.
“If we don’t, we’ll have failed,” he said. “Some companies will get it right and they are the ones who will outperform.”
Wherever you are on the AI journey, speed is crucial. He advised embracing the AI culture, accelerating its adoption, and making big bets on reinventing parts of the organization.
Dalman urged people to get hands-on with AI and work out what bits of the organization to focus on first.
With collaboration so important, Malik said knowing which partners will be the best fit for you is essential.
Baughan concluded that it’s refreshing that the ad industry is viewing AI optimistically, rather than fearfully — something which can only help push the debate forward.
For all of Pioneers | Conversations on our Industry’s Future see the hub.
Campaignliveus
