Orange Branding is a Dutch Brand Strategy and Design agency based in Amsterdam and operating in China (Shanghai, Guangzhou, and Shenzhen) since 2005. They have clients of various sizes in various industries, but in this project, we focused on companies interested in Cross-border e-commerce. When these businesses want to operate abroad, they will need a well-built brand. However, the value of branding has not yet been widely understood in China. Therefore, we ought to design a method to improve this understanding. Furthermore, research has shown that clients and brand designers need to work together to achieve the most optimal result. Accordingly, the research question answered in this thesis reads:
How can we improve co-creation in the brand development process between Dutch designers and Chinese clients?

To answer the research question, we have analyzed existing branding methods: Cultural Branding, Co-creating Brand Meaning, and OB’s usual methods. Furthermore, we researched Chinese culture in general and, regarding branding, to determine how a European designer should collaborate with a Chinese client. Lastly, we examined different Artificial Intelligence (AI) types to use in a new Branding Method.
What makes Cultural Branding unique is its story-based approach. Douglas Holt believes a brand has to offer a story, or myth, to its consumers. Consumers must want to use a brand’s products in order to live that story and build identity.
Co-creating Brand meaning tells us that the company behind a brand is the source of inspiration for the Brand Identity. Knowing this, a designer should not exclude the client from the brand development process but involve him in. However, the client must not acquire too much creative freedom, so it is the designer’s job to balance involvement and creation. We argue that AI can provide tools with which a designer can achieve this balance.
We conducted a pilot experiment to test the value of AI as a tool in brand development. Since the client inspires the Brand Identity, we wanted to see if we can use Machine Learning (a type of AI) to determine brand personalities. Therefore, we collected approximately 3500 images, labeled with one of Aaker’s Five Dimensions of Brand Personality: Sincerity, Excitement, Competence, Sophistication, and Ruggedness. Therefore, we created the Orange Branding ‘Aaker Personality Prediction Algorithm’ or ‘OB-APPA’. APPA is an ML algorithm that was trained using these images and is able to predict the personality of an image.
The OB team in China posed as imaginary clients and supplied new images for six fictional brands. Then, with the predictions from APPA, we manifested these predictions into visual brand styles. Next, the OB China team provided feedback on whether the created styles fit their intended fictional brand’s identity.
After examining the feedback, we conclude that the use of APPA could be very beneficial for a design agency like OB. OB can benefit from using APPA as their clients often supply a design brief which leaves almost complete creative freedom. APPA could efficiently deduce the desired brand personality, which will narrow the design brief. Consequently, an additional step in APPA could also predict the associated Brand Archetype. Combining the improved APPA with generative branding tools like Tailor Brands, IDEO’s font map, Front brief, and Colour Thief can generate branding elements in a short amount of time and make the system usable from anywhere.