“How to keep your brand consistent when AI tools can generate infinite variations?” This critical question was recently posted by Paul Dervan, Miro’s Head of Brand Marketing (1). He went on to share a brilliant example of how mobile network brand O2 has figured this out. O2 worked in partnership with VCCP’s AI creative agency, faith, to develop new AI-enabled tools to drive consistency in how they amplify distinctive brand assets (DBAs). Specifically, VCCP have helped O2 create fresh executions of their brand mascot, Bubl, that are consistent over time and over multiple comms channels.
In this post, we explore how VCCP have helped O2 harness AI to create “fresh consistency” and some of the key learnings.
About O2’s DBA: Bubl
The first thing to highlight is the importance of creating a distinctive brand asset in the first place. And then committing to the principle of amplifying it consistently, both over time and over channels. In the case of O2, the brand has been consistently using their Bubl character for over four years. “AI isn’t creating something new, it’s amplifying something that already connects with people,” as Paul rightly points out (1).
Bubl was launched in 2020 to “be the embodiment of the O2 brand”, as VCCP explain (2). “Bubl takes the form of a loyal and reliable blue robot helping ensure customers have the best possible service. Just like the network, Bubl is there whenever needed.”
The AI problem …
In the hands of marketers trying to “DIY” imagery using AI, the results can be hit & miss, as we posted on recently here. Using the example of O2’s Bubl character, VCCP show in a video case study how attempts to create imagery with AI ran the risk of destroying brand consistency (see below).
The AI solution
VCCP’s AI creative agency faith built a bespoke platform for O2: The Bubl Generator. This tool creates imagery of O2’s brand character Bubl based on client prompts (see below). “The character can be used in channels where cost or time constraints previously limited its use,” VCCP explain (3). “This allows for quick, reactive work across channels including social, web, apps and internal comms.”
The AI learning
The key learning for me, on reading more about this case study, is the time and effort invested to get the AI-enabled tool right. “faith trained the Gen-AI model to create images of Bubl in line with brand guidelines,” write VCCP on their website (3). This was not an easy, one-off job.
Multiple rounds of training were needed before the AI was able to consistently create high-quality and usable images. Furthermore, faith needed a 3D model of the Bubl character, so that the AI could be trained on poses of Bubl not seen before. “Previously, Bubl had only been seen from the front, which limited its use,” faith explain. “To remove these barriers and enable images to be generated of any potential scenario, further design work and modelling was necessary.”
In conclusion, to quote again from Paul, O2 and VCCP have shown how to “use AI to scale a distinctive brand asset across more touchpoints, faster. Not replacing the original, just making it work harder.”
Hats off to O2 and VCCP/faith for a brilliant example of how to harness AI.ing
SOURCES
