Coca-Cola
AI Brand Partner
Coca-Cola has engaged VML to create an AI brand partner tool as a Chrome extension. This tool will provide real-time feedback to Coca-Cola employees, assisting in the creation of marketing materials and briefs that adhere to the brand’s visual and tone of voice guidelines.
By serving as brand “guardrails”, the tool will correct any deviations from these guidelines, ensuring the production of consistent, on-brand content.
Company – VML
Role – Lead UX Designer
Year – 2024
01 Challenge
Individuals responsible for producing marketing materials often face challenges in adhering to brand guidelines, particularly when working across diverse markets. Cultural nuances can be overlooked, leading to briefs that lose their intended meaning. This inconsistency creates pressure on senior management, as they are required to manually review content for adherence to branding and cultural appropriateness.
Hi-fidelity mockup
02 Research
We conducted user interviews with various stakeholders and people in the business to understand their workflows and tools. The insights gained will inform our design decisions, ensuring the tool we create integrates seamlessly with their existing tools, such as Microsoft Word and PowerPoint online.
03 Goal
To simplify users’ existing workflows and avoid creating additional barriers or workloads when developing marketing materials and briefs, the tool must be designed with user experience in mind. Furthermore, it should not be intrusive; instead, it should assist users as needed by employing a sophisticated language model (LLM). This model should be carefully trained to ensure that users remain focused on delivering excellent results, supported by cross-referenced data and information from multiple sources.
04 Design
The tool was deliberately designed to seamlessly integrate with the user’s existing workflow, ensuring it does not disrupt their tasks but instead assists when requested. The streamlined user interface enables users to concentrate on their immediate tasks while delivering valuable insights and support at the click of a button.
By utilising a well-established and highly adaptable design system, high-fidelity mockups were created and tested with users. Concurrently, the engineering team developed the user interface and a fully operational language model, which underwent further testing to yield more meaningful results and insights from users.
Proposed design
