The industry does not lack for conferences. What it often lacks is a room where conventional thinking gets tested rather than confirmed. That was the premise behind Media Spotlight.
This year brought three speakers to a single theme: The Nuance of Navigation. The question was not whether the industry needs to change, but what it should keep, what it should adapt, and what it should stop mistaking for effectiveness.
With around $3 billion of media investment moving through Aotearoa each year, those decisions matter.
Felipe Thomaz, Associate Professor of Marketing and Director of Research at Oxford's Future of Marketing Initiative, arrived with data. A lot of it: 1,084 campaigns, 557 global brands, 51 countries, and ten years of results.
His central finding was uncomfortable for anyone optimising primarily for reach. Word of mouth reaches 16% of audiences but influences purchase behaviour in 48–72% of cases. TV reaches 57% but influences somewhere between 22% and 57%. The gap between exposure and behavioural impact is wider, and more variable, than most planning frameworks assume.
The performance numbers compound this. The average campaign in the dataset could be 2.6 times more effective with better channel construction. The most commonly used campaign archetype ranked second worst for outcomes. And "everything, everywhere, always on" was, in Thomaz's words, "absolutely the worst."
Channels are not interchangeable. They affect consumers differently, perform differently at different stages of decision-making, and interact with each other in ways that reach metrics do not capture. Adding spend without understanding those dynamics does not automatically scale effectiveness. It can work against it.
His conclusion: "Scale and availability without brand differentiation equals long-term economic peril."
Reach is a useful input. It is not a reliable measure of whether a campaign is working.
Eugene Healey argued that many of the systems brands use to manage communications were built for a media environment that no longer exists.
Linear campaigns. Layered approvals. Static brand guidelines enforced from the centre. That model made sense when audiences were concentrated, channel mixes were simple, and a single execution could reach most of the market.
Eugene argued that environment is gone.
TV's share of viewership among three major networks has fallen from 70% to under 30% since 1985. In Australia, more people now watch YouTube weekly than broadcast TV. Algorithmic distribution means brands no longer control the order in which their communications reach people.
As Healey put it, "positioning does not register as a story, or a series of messages. It registers as a vibe."
The governance model breaks at volume. Healey referenced one company where the CEO approved each radio ad individually. That process cannot function when dozens of assets need to go to market daily. The result is either a bottleneck, too little gets made, the brand does not register, or an incoherent volume of content that builds no consistent associations.
Healey's reframe was precise: "Your brand isn't any single execution. It is not the positioning that sits in the middle. It's the system that enables pattern recognition in a situation where there are potentially infinite autonomous executions."
Coherence, not control. That is the more durable response.
Melissa Clark-Reynolds ONZM presented AI as a strategic and cultural opportunity for our industry.
Her central argument: AI is a capability multiplier, not a headcount reducer. Tool adoption without frameworks creates chaos. Volume, in an environment already saturated with generated content, is not the win.
As synthetic content increases, the harder-to-automate advantages become more valuable. Strategy. Taste. Editorial judgement. Relationships. Cultural understanding.
"AI doesn't understand culture. You do."
She also reframed how teams should think about working with AI. "You are not prompting. You are briefing." The quality of the output is shaped by the quality of the thinking behind it.
That is not new. AI just makes it harder to ignore.
The implication is not that human expertise becomes less important as automation increases. It becomes more important. The more commodified production becomes, the more valuable judgement becomes.
The three arguments were connected.
Thomaz showed that reach optimisation without understanding how channels actually influence behaviour produces campaigns that measurably underperform. Healey showed that the governance and production models many brands rely on were built for a media environment that has fundamentally changed. Clark-Reynolds argued that as AI increases the volume of everything, the value of human judgement, taste, and relationships increases with it.
None of this requires abandoning established practice. The case across all three sessions was more specific: some fundamentals remain sound, and some assumptions underneath them do not.
Knowing the difference is the work.
Media Spotlight 2026 was hosted by the Commercial Communications Council at Q Theatre, Auckland, and sponsored by MediaWorks.