As more challenges are added to marketers’ plates as a result of AI, from synthetic data overload to increasingly fragmented audiences, industry leaders are warning brands against chasing endless audience segments at the expense of understanding who their real consumers are.
Speaking at a breakfast hosted by ad tech company Ogury, in partnership with B&T, titled ‘The Decisions That Will Define Marketing’s Next Five Years’, the panelists from Ogury, Havas, The Iconic and The Australian Centre for AI in Marketing (ACAM) argued the importance of using reliable audience information, questioning the value of new data points and focusing on insights they know will translate into meaningful action.
Showcased at the breakfast was Ogury’s own Persona Intelligence technology, designed to help marketers overcome signal loss and bridge the gap between audience planning and campaign execution. Rather than tracking individual users, the technology groups audiences into anonymised personas based on shared behaviours, interests and purchase intentions, helping brands identify and reach the audiences that matter most.
Chris Manson, country director ANZ at Ogury, Molly Blacker, national head of strategy at Havas, Louise Cummins, CEO and co-founder of ACAM, and Vonne Lombard, director of brand and Market Engagement at The Iconic, explored how marketers can make better decisions as AI creates new opportunities to generate data, while audience fragmentation makes reaching the right people increasingly complex.
‘Don’t drop the glass balls’
Havas national head of strategy Molly Blacker warned marketers against getting caught up in the constant pursuit of new data points and audience segments, arguing that knowing what to ignore is “just as important as knowing what to pursue”.
“Someone once told me ‘You’ve got a whole bunch of glass balls, and you’ve got a whole bunch of plastic balls, and the aim of the game is don’t drop as many glass balls as possible’.
“That may be a home thing or a work thing, but in marketing it’s about whether you have a common language around what those glass balls are? If you keep catching every single plastic ball, every single data point, you will drop glass balls,” she added.
She said marketers chasing an audience in an age bracket such as “25 to 54” should not think their audience’s age is the most interesting thing about them.
“When you think about all the times you have connected with someone, was it about their age?” she asked.
“That’s actually the least interesting thing about us, so I think it’s about finding the human part of your audience and knowing ‘That’s the glass ball’ and if they start suddenly exhibiting this behaviour then that’s a glass ball for us.”
Ogury country director ANZ Chris Manson agreed that marketers need to get a clearer understanding of their priority audiences before deciding which channels to use to reach them.
“We’ll often get briefs that are pretty basic on the audience expectations,” he said.
“We’ll get a home loan brief, for example, first home buyer brief saying this is exactly what they look like, this is their age, and they’ve got this much money to save, and go get them.”
Manson said consumers are more complex than these narrow audience definitions suggest, with customer journeys extending well beyond the obvious websites and channels associated with a particular purchase.
“Consumers are always multidimensional. They’ve got so many layers behind them, particularly for the customer journeys that they’re going on,” he said.
“It’s not as fixated as they’re looking at these property websites, and then they’re going to go to these comparison websites, and that’s the ecosystem that they stay in, they don’t fall outside of that at all.”
Focusing exclusively on those expected touchpoints risks overlooking other environments and opportunities to reach potential customers, Manson said.
“If you’re focusing just in that siloed ecosystem, then you’re missing out on so many other touchpoints and environments and communication areas that you can start to focus on to then bring that scale without then sacrificing accuracy at all.”
He also encouraged marketers to lean on their agency and technology partners to help make sense of the vast amount of audience data available.
“Lean on your partners, lean on your agencies, because like we said before, there’s so much data out there,” he said.
“But a lot of those agencies and a lot of the vendors, like us being one of them, have been able to really condense and make that data a lot simpler to understand and to pinpoint who those most priority, those greatest priority audiences are.”
‘We’ve not moved anything, we’ve just thrown it around’
Manson added that the sheer volume of information and technology now available to marketers was creating “a growing challenge for the industry”.
“There’s too much stuff. We’ve got way too many things that we have to now focus on. We’ve got too many signals, too much tech, too many options, too many decisions,” he said.
As AI makes synthetic data and audience insights easier to generate, marketers face an increasing number of signals competing for their attention. But more information does not necessarily mean a better understanding of real consumers.
“Real information to people is still going to be the most effective means to understand audiences,” Manson said.
He also urged marketers to ask tougher questions of the technology vendors supplying their data and insights, rather than accepting claims about their capabilities without scrutiny.
“No vendor, no tech solution out there shouldn’t be able to tell you how they are specifically doing what they’re doing,” he said. “If it’s hidden, then there’s something hidden for a reason.”
THE ICONIC’s director of brand and market engagement Yvonne Lombard acknowledged that while a big challenge is the amount of data available to businesses, it’s more so the tendency to chase it without understanding whether it will make a meaningful difference.
“We’ve got paralysis,” she said. “There’s a lot of information, and we don’t know if it’s actually the big shifter or big needle mover. And then we run like dogs after the small data points that we don’t necessarily know what the impact is,” she said.
Lombard compared the process to a dog chasing a ball in a park, with teams repeatedly running after individual data points without making meaningful progress.
“So we get the data point, we throw it, the team runs behind, fetch the data point, bring it back, and then they sit and wait. And then you throw the data points, and then they run after it, and they come back, and they wait,” she said.
“And what we’ve done is we’ve actually not moved anything; we’ve just thrown all around. And that is a big challenge that we have in marketing, where we have so much data, and it talks about the cognitive overload, or choice overload, or paralysis that the team’s really struggling with.”
Lombard said the volume of available data can also lead businesses to apply the wrong measurement principles to the wrong objectives, particularly when balancing long-term brand building with short-term performance targets.
“We want long-term brand building approach, but we’re actually managing it like it’s a performance marketing engine, and we constantly have to manage and navigate that tension within marketing,” she said.
That tension can become particularly clear when businesses commit to long-term brand investment but face pressure to deliver immediate quarterly results.
“We say we need to get to the 60/40. We need to invest to the brand long term. And then you get a question: Is okay, but if you don’t, you know, get the quarter’s growth, or we don’t have this type of results, then we’ll pull back on the brand budget,” Lombard said.
She said a stronger decision-making framework and a shared understanding of what different measurement tools can — and cannot — tell marketers would help businesses use their data more effectively.
“If we have a common understanding throughout the business, I think it would be a lot easier for us to use our data better,” she said.
“It’s not about using the data; it’s actually moving the data to insights, implication, and an action plan.”
How should AI be used?
Australian Centre of AI in Marketing (ACAM) CEO and co-founder Louise Cummins, who has worked in marketing for 30 years, said the answer to information overload is not simply adding more AI tools, but ensuring the data and systems behind them are fit for purpose.
“Sometimes there’s just this inundation, and then it creates inertia. And the marketers just think, ‘I don’t know where to go’,” Cummins said.
“The challenge is that AI shines a big light on poor information architecture,” she said, warning that introducing AI without addressing inconsistent data and weak foundations risks amplifying existing problems rather than solving them.
Cummins said marketers need to ensure they are feeding AI the right information, weighting the right signals and being honest about what the data actually shows.
“Marketers are very good at reporting when things work, and not so good when things don’t work,” she said.
She highlighted use cases of businesses exploring AI agents that can draw information from different sources and help marketers make better-informed decisions, and doing it well.
“One particular client is actually looking at creating an agent so that every time somebody sends an email, you can actually say, does that actually connect with what the market signals and what’s happening in the market?” Cummins said.
Another client, she said, is exploring ways to interact directly with its data rather than relying on dashboards.
Cummins added that getting the information architecture right also changes how much supervision AI needs. While newer models may be able to operate more independently, she said human oversight “remains essential” – particularly when major decisions are involved.
“AI should always just be a tool. If you’re leading with only AI, then it’s the wrong choice,” she said.
