Rodney Greenfield: "Human Intelligence Plus Artificial Intelligence" Elevates Creativity

Rodney Greenfield is a seasoned technology leader with over 26 years of experience delivering data-led transformation. At AKQA, he has focused on the growth of the data and insights practice, spanning machine learning, data engineering, MLOps, and advanced analytics. Rodney brings deep expertise in applying AI to customer experience research and optimisation, including the development of conversational and agentic AI solutions.

His work centres on turning complex customer, behavioural, and operational data into actionable insight, enabling more effective decisioning, personalisation, and end-to-end customer experience design across web, mobile, and emerging conversational channels.

Rodney> The question is coalescing around two things: “What are AI systems saying about our brand, and how do we influence that?” and “What can culture, search and social demand tell us about where growth is moving next?” Clients are not just asking for reporting; they are asking for a better radar, better levers, and a clearer view of where experience, content, and commercial growth intersect.

Rodney> The answer is HI + AI: human intelligence plus artificial intelligence. AI is useful as a sparring partner, a challenger and a way to interrogate evidence, but the more homogeneous AI-generated work becomes, the more valuable distinctive human taste, judgement and creative curation become. The next frontier is not letting AI flatten the work, but training it around codified stylistic decisions, brand behaviours and creative principles that protect distinctiveness.

Rodney> One strong example is using three-year longitudinal social impact analysis we did for a mobile phone brand. It was across product conversations to understand not just volume, but shifts in perception, marketing effectiveness and brand meaning over time. Grouping by product sets, that kind of analysis can reveal what each generation changed in the market’s mind, not just what people clicked on. We can then use codified synthetic populations to test early creative and design directions before release, helping teams understand likely audience response while there is still time to shape the work.

Rodney> The question is not “should we have first-party data?” It is “what decisions will this improve?” If it helps you increase CSAT, improve brand perception, lift engagement, reduce friction, grow loyalty or make smarter investment decisions, then it matters. A first-party data practice should teach you something useful about your customers, partners, products or brand perception, not just create another owned dataset.

Rodney> Data is often used for lazy optimisation: what will affect the most experiences, most cheaply, with the least change. Used creatively, data should uncover new possibilities for individualisation, experience design, cultural participation and customer value. The power of personalisation is not just showing someone the next product; it is understanding the moment, the need, the context and emotional context well enough to make the experience feel genuinely useful.

Rodney> Every dataset is a lens, not the truth. Social data, search data, first-party data, AI answer visibility and campaign performance all carry bias, context and blind spots. The discipline is triangulation: look for patterns across sources, make assumptions visible, and separate what the data proves from what it merely suggests.

Rodney> Trust in data is not achieved by making dashboards look more authoritative. It comes from provenance, quality, consistency, explainability, and knowing whether the data is fit for the decision being made. The danger now is that AI can make weak or partial data feel polished, so trust needs to be designed into the workflow through source visibility, confidence, lineage and human review.

Rodney> You need a global spine and local intelligence. The global layer should define governance, consent, taxonomy, security, measurement, and responsible AI principles; the local layer needs people who understand market context, cultural nuance, regulatory expectations, language, behaviour, and the “so what”. That balance is becoming more important as regulation evolves, including Australia’s Privacy Act reforms around a Children’s Online Privacy Code and serious invasions of privacy, and the EU AI Act’s staged transparency obligations.

Rodney> A responsible data practice starts with purpose: why are we collecting this, what value does it create, and could it cause harm if used poorly? Practically, it means consent-aware collection, data minimisation, secure access, clear retention, quality checks, bias testing, explainability and accountability. Strategically, it means data should improve decisions and experiences, not just increase the organisation’s appetite to collect more.

Rodney> One misconception is that data is cold, mechanical or anti-creative, so people delay it until after the big decisions have already been made. The other is that AI is a magic wand: add sparkles, tap the problem, and the work becomes intelligent. In reality, data can expose the truth, and AI can accelerate thinking, but neither replaces judgement, strategy or taste (well not yet).

Rodney> The big one is democratised access to enterprise data through agents and MCP-style connections. MCP is explicitly designed to connect AI applications to external systems, data sources, tools and workflows, which creates huge opportunity but also raises the risk of people accessing data without the skill to interpret it correctly. The next battle is not just access; it is semantic governance, because even an operational semantic layer can present the right number in the wrong context, leading to confident but poor decisions.

Rodney> Social data tells you what the market is starting to care about; first-party data tells you whether your actual customers are behaving in ways that make it commercially meaningful. The opportunity is not to stitch everything back to individuals, but to connect cultural signals, audience demand and owned behavioural data at an aggregate, responsible level. That is where AKQA proprietary tooling (across social listening, answer intelligence, cultural intelligence, and growth modelling) become powerful together: one shows the signal, one shows the demand, one shows the customer behaviour, and one shows the likely value.