Here, Mary-Jay Hanna, account executive at Horizon Communication Group discusses a recent session which took place between Shanna Crispin, founder of strategic communications consultancy Measured Advisory, and government communication professionals. She discusses her key takeaways around how AI, behavioural science and human judgement are changing the way organisations understand their audiences.
Government communicators are no strangers to impossible briefs – shrinking timelines, squeezed budgets, fragmented audiences and the small matter of needing to shift actual human behaviour without setting off a stakeholder emergency alarm.
Recently, government communication professionals from across Australia tuned in to a CPRA Government Special Interest Group discussion on synthetic research and what it means for the future of communication strategy.
Facilitated by Shanna Crispin the session brought together Salem Lassoued, founder of Muja, and Dr Lisa Portolan, managing director of Horizon Communication Group, to explore how AI, behavioural science and human judgement are reshaping the way organisations understand audiences.
Understanding synthetic research
Lassoued explained that synthetic research uses AI to simulate traditional research methods, including focus groups, surveys, and audience testing. In practice, it allows organisations to explore how different audience segments may respond to messages, ideas, and communication strategies before committing to larger-scale fieldwork.
But this is not just research with a stopwatch.
“Synthetic research can really help government as a way to ask bigger questions, not just faster ones,” Lassoued said.
For government communicators working within tight timelines, stretched budgets and complex approvals, the appeal is clear. Synthetic research can help test early ideas, identify blind spots, and explore audience attitudes before a campaign direction is locked in and everyone agrees to pretend the first idea was always the best one.
Why audience understanding matters
Drawing on her experience leading communications and behavioural insight programs, Dr Portolan highlighted the role audience understanding plays in successful campaigns.
“Audience insight… is the lifeblood of a campaign,” she said.
That insight shapes everything from audience demographics and behavioural motivators to message development, creative testing, channel selection and campaign execution. In a communications landscape where audiences are scattered across TikTok, Reddit, Instagram, traditional media and at least three group chats before breakfast, assumptions can get expensive quickly.
A complement, not a replacement
The discussion returned often to the idea that synthetic research should complement traditional methods, rather than replace them. Lassoued said it can be valuable for rapid idea testing, audience segmentation, scenario planning, and message refinement before investing in fieldwork.
“It’s not about removing human budget, it’s not about replacing traditional market research,” Lassoued said. “It’s about being able to look at the deficiencies of traditional market research and use synthetic as a way to boost that.”
Both speakers were clear traditional research remains essential where representative samples, public opinion measurement, evaluation studies, policy testing, or evidentiary weight are required.
The same caution applies to community consultation. Synthetic research can help shape better questions, but it cannot replace genuine engagement with real communities. As Portolan noted, consultation and co-design are also about building trust, buy-in and relationships, all very much still in the human column.
Keeping humans in the loop
That human element was a major thread throughout the discussion. Portolan emphasised that while AI-enabled tools can identify patterns and generate insight, human researchers and communicators remain essential for interpreting nuance, understanding emotional responses and navigating stakeholder complexity.
“You can never really remove the human component,” she said, adding that even synthetic platforms rely on the right questions, prompts and research structure to produce useful outcomes.
Audience questions focused on accuracy, validation, data sources, and ethical considerations, reflecting growing interest in how AI can be responsibly integrated into government communications without everyone losing the plot, or the procurement team.
Lassoued pointed to “decision alignment” as one way Muja assesses reliability, looking at whether synthetic research leads to the same decision as traditional human research. He said Muja’s average sits “around 89 to 92 per cent decision alignment”.
Both speakers agreed the future is unlikely to be a simple case of AI replacing traditional research. Instead, Portolan suggested the industry will see greater integration of synthetic research into existing research practices, particularly as communicators look for faster and more flexible ways to test ideas and understand audiences.
She also noted that the broader debate around AI needs more nuance.
“The discussions that we’re having about AI aren’t particularly nuanced,” Dr Portolan said. “It’s about getting to those more interesting, nuanced, layered discussions about the possibilities that exist and the ethics.”
For government communicators, it’s clear that synthetic research is not a silver bullet. Used well, however, it could help teams sharpen strategy, stretch research budgets, and better understand the people they are trying to reach.
And in a sector where the brief is urgent, the audience is complex and the budget has already been asked to perform minor miracles, that is an opportunity worth paying attention to.
