
As algorithms reshape digital advertising around Creative Diversity, we helped Grab move early to stay at the forefront and maximise campaign effectiveness. As its preferred agency partner of 7+ years, we brought proprietary AI and creative automation together to turn this new model into a scalable regional advantage, evolving in step with Grab’s ‘AI-First with Heart’ philosophy.


Digital ad campaigns used to scale one master creative into personalised variations that largely looked alike. Now, a new algorithmic model has flipped the script. It calls for multiple distinct ideas and visual concepts to identify which creative resonates with which audiences. For Grab, this meant moving beyond adaptation to produce genuinely different creative routes at speed and scale, without compromising brand consistency or local relevance.


To put Grab at the forefront of this shift, we built BrandGPT, a proprietary in-house AI model designed to generate and develop genuinely differentiated creative concepts. It combines three layers of Grab-specific intelligence:
Trend analysis turns online conversations and emerging trends into timely creative opportunities.
Sei-katsu-sha uses simulated audience personas to uncover human-centric angles during ideation.
Brand’s ‘brain’ learns from Grab’s guidelines, past creative work, and campaign performance data to shape ideas that fit the brand.

Grab Philippines led the first rollout across GrabCar, GrabFood and GrabMart. The campaign explored different consumer moments and motivations, from late-night rides and cravings to grocery restocks and party needs. BrandGPT helped define each creative route and its visual style, expressed through a mix of lifestyle photography, food-led imagery, character-driven scenes and 3D illustrations. Our signature Digital Creative Automation (DCA) system then modularised, localised and rapidly scaled each route across formats and platforms.


Within two months, Grab Philippines’ campaign showed strong early engagement across all three platforms, with CTR significantly outperforming platform averages as the system continued learning and optimising the assets.
CTR on Google
CTR on Meta
CTR on TikTok

