AI Beyond Efficiency

From my conversations with CMOs and digital marketing leaders, there is undoubtedly an instinctive focus on AI for cost savings. However, this efficiency-first mindset risks underdelivering and whilst cost cuts are tangible, they overlook AI’s potential.

Despite cost-cutting being one obvious approach to AI use cases, the non-obvious powerhouse behind this technology is value amplification: building commercially viable solutions that create additional income streams or ‘enrich’ existing ones.

I want to put forward a new challenge to marketers: embrace AI as a co-creator for transformative growth, drawing from real campaigns to illustrate how ideas that combine the best of human + AI can reinvent experiences and drive revenue.

AI's Reactive Nature: The Human Spark That Ignites Breakthroughs

My inspiration for proposing this challenge deepened upon realising AI's limits: it is reactive and needs human input to activate, even in agentic setups, someone needs to press the ‘go’ button. This highlights our edge: cultural nuance, empathy, risk-taking, and unexpected connections, which fuel non-obvious innovations.

Take the Iron Man suit: Jarvis executed but Tony Stark envisioned. Serendipity drives many breakthroughs, as seen in scientific discoveries like penicillin. For brands, this means using AI to scale human ideas into viable offerings.

Ditching the Binary: Efficiency and Amplification for Holistic Impact

We must move beyond a binary view: don't choose cost cuts or output boosts, do both. Free resources from repetition to innovate, creating amplified value like new AI-powered services atop existing products.

McKinsey notes generative AI could add 5-15% to marketing productivity, while agentic AI might automate 45% of tasks by 2027, unlocking time for revenue-generating innovations.

I am aware that within larger corporate ecosystems, transformational change across the adoption of technologies can be tricky - people can resist behavioural change. Yes, infrastructure requires adjustments, but we should aim to empower and adapt as a priority.

Real-World Fusion: AI-Powered Campaigns That Drive Cultural Shifts

Inspired by these possibilities, I want to share three separate (and very different) campaigns blending efficiency with amplification for cultural and commercial wins, proving AI can spawn new income streams.

To promote their extra-smooth new product, Heineken Vietnam rewarded fans during live football matches. For every goal celebrated with a player sliding smoothly, Heineken rewarded fans with discounted products delivered instantly by Grab. A custom machine learning model and live player data measured the slides, with longer slides providing a bigger discount. It’s the perfect example of fusing human excitement and unpredictability with tech precision. The result? More than a 34x sales increase during the first match and 30,000 litres sold in a two week launch period. Not just revenue spikes, but a new engagement model.

A large insurance provider in Southeast Asia used Agentic AI to turn complex PDFs into nuanced videos, facilitating faster decision-making, 80% production efficiency, and significant visibility boosts, which gave way to cross-platform upsells as fresh revenue.

Creativity is filling gaps of ambiguity, and this case study exemplifies this wonderfully. As part of its sponsorship of the China Hong Kong Paralympics Committee, Cathay Pacific wanted to shine a light on the undocumented history of past Hong Kong Paralympians. Using AI, they transformed voice recordings and pictures into a video celebrating the 100+ medals that the Paralympians have won over the years. The video launched during the opening ceremony and rewrote these winners into digital history, filling gaps that humans couldn't and increasing overall awareness and engagement. This example highlights how ethical amplification can be used for brand affinity.

Building the Future: Agentic Ecosystems for Limitless Creativity

My core belief is that AI's less obvious value lies in amplifying base solutions. A great example of this is agentic AI: continuous agent collaborations fed by data inputs and instructed outputs for ideation and creation.

The hurdle? Not data - we’ve got tons of it, but the orchestration of this data into outcomes like AI-driven personalisation add-ons. Future creativity lies in the ‘mix’ of data inputs, the instructions of what to do with these inputs and the outputs that agentic processes can produce.

McKinsey sees agentic AI powering workflows and potentially boosting productivity 3-5% annually. This enables brands to layer innovative services, turning efficiency gains into amplified income. Conversely, Gartner, in its study with MIT Media Lab and Project Nanda, warns 95% of AI pilots fail due to a lack of clear strategy or business outcomes, poor data quality and governance, organisational resistance, integration challenges and more. It’s crucial to note that failures are often a result of an overemphasised focus on superficial efficiency in sales and marketing pilots, which consume 50-70% of all AI budgets.

By prioritising human-AI hybrid models that blend efficiency with amplification, we build sustainable value and increase the likelihood of success.

Scaling AI Responsibly: Collaboration, Empowerment and iteration

To realise AI's full potential, we must foster cross-division collaboration, aligning procurement and compliance with marketing and operations to explore use cases and ensure safety, both in the technologies’ use and outputs.

I urge leaders to empower individuals beyond basic experimental setups: Create safe spaces for hands-on trials with new tech, then amplify learnings and successes organisation-wide. Many companies have built out AI councils made up of different competencies or division representatives that can help disseminate information more quickly.

Finally, start small; isolate experiments to specific tasks or divisions, avoiding company-wide rollouts initially to iterate effectively and minimise risks. Through this approach, success is scalable, whilst experimental mishaps and their impact can be minimised.