Branding is often considered a highly imaginative and artistic exercise. However, it is more of a data-driven process based on an in-depth analysis of a brand's qualitative and quantitative data.

Since performing this analysis manually could be incredibly time-consuming, little room is often left for the truly creative aspect of branding. On the other hand, being too brief on the analysis and relying solely on creativity may cause the brand to deviate from its intended vision, mission, or positioning.

That's where AI comes into the picture. Artificial Intelligence can assist us in processing large amounts of data that are impossible to analyze alone while respecting project timelines. However, it is crucial to remain vigilant to avoid biases that could sneak into the process both from AI and ourselves.

Can AI help us speed up the analysis processing? How do we ensure not to let ourselves be influenced by its biases or influence it with ours?

‍

We must shift our perspective on the creative process when working with AI. To achieve high-quality results, we need to implement certain behaviors, such as using diverse and high-quality datasets, verifying the accuracy and validity of the AI model, and integrating human control during critical decision-making moments to avoid bias and make responsible design decisions.

When properly implemented with these necessary precautions and checkpoints, Artificial Intelligence does not replace designers' efforts but instead supports them in enhancing and accelerating the work processes. Such aid enables us to significantly expand the range of outcomes in our work during both the divergence and convergence phases in ways that would otherwise be unfeasible due to limitations in time and resources.

‍

Our process is based on the branding methodology developed by Amanda Louisi, a senior brand designer specializing in personal and corporate branding. This methodology emphasizes the importance of building a solid foundation by starting from data that will provide clear guidelines and boundaries once entering the creative part.

Amanda Louisi's methodology

The first step is to gather and analyze as much quantitative and qualitative data as possible on the brand to extract keywords that can help her determine the brand's personality traits. Then, she create a catalog of shapes, colors, fonts, and symbols that effectively convey the brand's core values and characteristics inspired by the brand's personality. Finally, the creative process can begin on this very well-defined canvas. Amanda applies and re-elaborates the catalog's visual elements to the typical artifacts of the brand identity, building a cohesive and strong identity.

The guidelines and limitations she established during the process help her to ensure the deliverables met the brand's true communication needs.

Buildo’s methodology

Inspired by Amanda Louisi's method and the design thinking principles, we initially created a branding process that was something like this:

During the initial phase of divergence and convergence, our goal is to identify the core values of the brand rather than its personality traits. To achieve this, we thoroughly analyze all relevant documents related to the product. We may also collect data ourselves through interviews and questionnaires. In the second phase, we build the catalog of visual elements from the basis of the values as suggested by Amanda.

The analysis process was completed using analog methods, typically consuming 70-80% of the total project time and leaving limited time for the creative part of the brand identity.

Along came AI.

As a company that gathers curious and innovative minds, Buildo strongly believes in letting its people experiment with AI's potential to understand how it can enhance processes and growth productivity. In this context, AI is never seen as a replacement for human capabilities but as a tool that can augment and complement human skills, freeing up time for non-AI tasks.

So, recently, our branding service has undergone a significant transformation thanks to the incorporation of AI technology in the process.

With the help of Artificial Intelligence, we can now broaden the horizons of the divergence phases enormously. However, this exponential increase in brainstormed ideas does not worry us since, in the same way, AI can make us converge rapidly, even more quickly than before.

As a result, we stopped describing the process as a double diamond but a Double Cusp.

But let’s see the process steps in detail, retracing the phases of VisionSphere’s rebrand.

VisionSphere’s rebrand

When Vehnta contacted us, the company requested that VisionSphere have its own brand identity instead of using the company's image. During the project kick-off, we thoroughly investigated the communication needs and discussed the Vehnta team's ideas for the product's look and feel.

We analyzed VisionSphere's existing brand, website, application, visuals, and copy. Then, we identified inconsistencies and communication elements that differed from Vehnta's vision, mission, and goals for the product.

We converted all the information we collected into text, using direct quotes and avoiding reinterpretation. Next, we asked ChatGPT to analyze and extract keywords during multiple iterations. We specified system and user prompts and used different versions for each iteration to ensure accuracy.

After the first stage of divergence, we move onto a convergence phase, where we work again with ChatGPT to experiment with various combinations and clusters of keywords. The main objective is to identify three to five core values that effectively summarize the brand's unique characteristics.

This is also the first checkpoint where the generative work of ChatGPT meets the control and challenge of the designer, who, thanks to their experience and know-how, verifies the quality and goodness of the output.

For VisionSphere, the three values that emerged were Innovation, Accuracy, and Reliability.

After establishing our brand values, it was time to diverge again and explore all the visual possibilities we had to communicate them. We asked ChatGPT what the best shapes, colors, fonts, and symbols would effectively convey the brand values to VisionSphere's target audience. In the same way, we asked for a list of any visual elements that might hinder our communication efforts so that we could avoid them. As in the previous divergent step, the more numerous and broad scenarios we tested, the better.

Once again, ChatGPT helps us to come together quickly with a cohesive list of visual elements that might or might not be used to convey brand values effectively. It is now up to the designer to evaluate the advice provided by ChatGPT and compare it with the brand's requirements and the client's feedback.

This will enable us to effectively communicate all the nuanced meanings and prioritize the development of a clear and comprehensive messaging strategy.

Our ultimate goal is to create consistent brand communication that enables our clients and stakeholders to internalize the brand voice and communicate its values easily.

The main change in the rebranding process was the brand's color, which was switched from green to violet. This change is because an essential aspect of VisionSphere is its geographical and territorial approach, visually represented through maps.

However, when the map iconography is associated with the color green, it tends to bring ecology and the environment to mind, which differs from what VisionSphere intends to communicate. Therefore, we decided to use violet, which is more relevant to the brand's values and the technology and data theme.

Thanks to the consistent and incremental work done to establish the foundation of the brand identity, we efficiently created the interfaces for both the VisionSphere website and the application in a short timeframe, maintaining the highest quality and attention to detail we require from our outputs.

Elide each other biases

One of the most common challenges when integrating ChatGPT and other AI tools into design processes is the biases that algorithms inherit from their training dataset. However, this fear should not stop us from employing the immense power and possibilities that come with generative AI tools. If we think about it, as designers, we are called daily to recognize and monitor our biases. We ourselves often fall into the trap of looking for patterns that confirm our hypotheses and ignore those that contradict them.

Artificial Intelligence is incredibly useful in processing large amounts of data that would take a team of designers a long time to analyze, exponentially broadening our horizon and often proposing unexpected and uncharted solutions.

From our side, we designers can use our experience and know-how to refine further and improve the output generated by AI, giving it a human touch and ensuring its quality.

The Designer-AI duo is thus able to compensate for each other's weaknesses and biases, making the most of their strengths and skills.

When discussing the role of AI in the realm of design and graphics, preconceived notions often arise that are linked to the quality of the work. These prejudices often give the impression that the entire creative process is solely delegated to the machine, which can result in doubts about the final product's authenticity, originality, and quality.

Based on our experience, the best approach to utilizing AI in creative processes is to delegate the tasks it is proficient at, such as quickly processing large amounts of data multiple times and leaving the designer plenty of time to do what they do best: design.

Since we had the chance to evaluate this unique combination thoroughly, our experience not only the production of high-quality output that is cost-effective and aligned with the needs and preferences of our customers but also the significant enhancement of the overall quality of life of our designers, by embracing this approach, we have achieved outstanding results that have positively impacted both our business and the individuals who are part of it.