Leveraging AI to Build Innovative Brands: Use Case - Part 4 of 5

Leveraging AI to Build Innovative Brands: Use Case - Part 4 of 5

Considering a new brand to better represent our AI expertise, we plan to depart from our decade-old, generic ‘System in Motion’ name. We will use generative AI to suggest unique, protectable brand names.

In this series of five articles, we will attempt to build a new brand identity. We will see how much AI can help us, analyze mistakes in our prompts, and learn more about how to leverage AI in creative processes.

Disclaimer: we are not branding experts; we are exploring this field through our company’s journey. We don’t advocate replacing branding agencies. Many companies don’t have the budget or access to professional agencies. These companies can benefit from generative AI.

In the first and second articles of the series, we covered the basics of establishing a brand. When we started, we only had our Purpose, Mission Statement, and our Core Values. In six prompts we established, in great detail, all our new brand elements .

In our third article , we tried and failed to generate a new brand name. All brand names we came up with were taken and impossible to trademark.

Different Approach

The major issue we had previously was that all the generated brand names were using existing words and concepts to combine into a brand name. To be able to trademark a new brand and own the associated domain name, it must not use common words. We need a word that does not exist in any language. How do we do that?

Prompt Neologism

List the 5 key steps for crafting a neologism, based on clear brand elements, to a list of brand names to test.

Answer

Creating a neologism, or a new word, for a brand name can be an effective way to ensure uniqueness and memorability in the marketplace. Here are five key steps to craft a neologism based on clear brand elements, leading to a list of brand names to test:

Define Core Brand Elements:

• Identity: Understand the core identity of the brand. What does the brand stand for? What are its values, mission, and vision?
• Target Audience: Identify who the brand is speaking to. What are the demographics and psychographics of the target audience?
• Unique Selling Proposition (USP): Pinpoint what makes the brand different from competitors. What can the brand offer that no one else can?

Brainstorm Base Words:

• Start with keyword brainstorming. List words related to the brand’s industry, products, and core values.
• Consider the emotional response or imagery each word evokes. This can be crucial in how the brand is perceived.
• Use linguistic techniques such as affixation (adding prefixes or suffixes), compounding (combining whole words), and blending (mixing parts of words).

Create Neologisms:

• Experiment with combinations of the brainstormed words and linguistic techniques. Morph words together in new ways to create something unique yet meaningful.
• Ensure the new words are easy to pronounce and remember. This increases the likelihood of brand recall.
• Check for cultural sensitivity and appropriateness in different languages if the brand is or plans to be global.

Evaluate and Refine:

• Assess the created neologisms for alignment with brand elements. Do they reflect the identity, appeal to the target audience, and communicate the USP?
• Conduct internal feedback rounds within the team or with focus groups to gauge reactions and get initial impressions.
• Refine the list based on feedback and linguistic effectiveness, ensuring the names are not only unique but also viable and marketable.

Legal and Linguistic Testing:

• Perform a thorough trademark search for each potential brand name to ensure legal availability.
• Test the names in various real-world scenarios and across different media to see how they perform. This includes checking domain availability for digital presence.
• Finalize a shortlist based on these tests and prepare for market testing to see how potential customers respond to each name.

By following these steps, you can create a list of innovative and effective brand names that are rooted in the brand’s core elements and ready for further testing in the market.

What is Missing?

This is unfortunately the same process that was followed when we used Chain of Thought and Tree of Thought. So it seems that we have reached the limit of what our Large Language Model (LLM) can generate as brand name creation process. Trying different prompts did not lead to a different process, and we already know that this process leads to brand names that already exist.

Let’s try to challenge the model by sharing our experience with the methodology. As a follow-up to the previous prompt and answer we asked:

Prompt TriedThatAlready

We tried this process and it led to brand names that exists, because they used full words. Give a different process that will not use full words.

Answer

Creating a brand name using partial words or entirely new constructions can help avoid the issue of ending up with names that already exist. Here’s a revised process focusing on creating more abstract neologisms:

Identify Key Brand Attributes:

• Clearly define the essence of the brand including its vision, mission, and values.
• Understand the emotional tone the brand wishes to convey - innovative, trustworthy, energetic, etc.

Initial Word Fragmentation:

• Start with key terms related to your brand’s attributes and services.
• Break these words down into syllables and phonetic fragments. For example, if ’technology’ is a key term, break it into ’tech,’ ’nol,’ ‘ogy.’

Invent Through Phonetics:

• Play with the sounds and structures of language. Focus on phonetics, the way the fragments sound when spoken aloud.
• Create combinations of sounds that are not necessarily tied to existing words but are pleasant or striking to hear. For instance, combining ’nol’ from ’technology’ with ‘vate’ from ‘innovate’ to form ‘Nolvate.’

Abstract Creation and Refinement:

• Develop a list of these new, phonetically pleasing word fragments.
• Refine them by considering ease of pronunciation, memorability, and the ability to convey the brand�s emotional tone.
• Ensure they are short, catchy, and distinct to facilitate brand recall.

Contextual and Visual Testing:

• Place each potential name in various mock-up situations (e.g., website headers, product packaging, advertising) to evaluate visual and contextual effectiveness.
• Conduct informal focus groups or digital surveys to gather feedback on the names’ appeal, uniqueness, and relevance to the brand.

Legal Screening and Linguistic Validation:

• Perform a thorough trademark search for each potential name to ensure they are legally available.
• Check the linguistic implications of each name in the primary languages of your target market to avoid negative connotations or misunderstandings.

Final Selection:

• Choose the name that best meets all criteria: legal clearance, marketability, alignment with brand attributes, and positive audience reception.
• Secure the necessary trademarks and domain names to protect the brand identity.

This process focuses on creating entirely new word forms by manipulating sounds and syllables, which can help in crafting a unique and legally safe brand name that resonates with your target audience and embodies your brand’s essence.

Great Idea, but…

The idea of breaking words in syllables sounds promising. This will multiply the potential number of combinations, and create a true neologism, that has a higher probability of not exist in any language, available domain name, and no existing trademark. But somewhere in the middle, we see we won’t be able to rely entirely on AI.

The example ‘Nolvate’ is a very poor example. Because ’nol’ is not a syllable of ’technology’. It can be broken down in tech/techno/logy, but no human will break if down in tech/nol/ogy.

The second issue is the emphasis on sound. As far as we understand the way Large Language Models operate, they manipulate tokens, which are related to words, and are interconnected by meaning. The relationship between words, syllables, and sound, will probably not be very accurate.

Chain of Prompt

Because of the above limitation, we will have to break down the process in a Chain of Prompts, with a human check between each prompt. This process is necessary when touching areas where LLM is a risk of making a poor decision, will not perform well. Chains of Prompt are the best illustration of human AI collaboration. For a process that is well established, with template prompts, it becomes a rapid back and forth between the AI and the human.

Here is how we are going to structure our Chain of Prompt:

• 8A: Let the AI generate a list of 30 keywords related to our brand elements, extract the syllables that most represent each word.
• Human: we will pick and choose the syllables we think are most promising.
• 8B: Run a Tree of Thought, with the updated syllables list.

Prompt 8A

Generate a list of 30 keywords which represents our brand elements. For each keyword, extract one of a few syllables from this word, that we can use to create a neologism as our new brand name.

Answer

Sure, here are 30 keywords that represent your brand elements, along with extracted syllables that could be used to create a neologism for your new brand name:

• Empowerment - Pow
• AI - Ai
• Mastery - Mas
• Training - Trai
• Success - Suc
• Guidance - Gui
• Integration - Int
• Education - Edu
• Innovation - Inno
• Transformation - Trans
• Leadership - Lead
• Quality - Qual
• Trust - Tru
• Clarity - Clar
• Professional - Pro
• Insightful - Sight
• Authoritative - Auth
• High-tech - Tech
• Impact - Pact
• Future - Fut
• Business - Bus
• Keynotes - Key
• Detailed - Det
• Valuable - Val
• Content - Con
• Accelerator - Accel
• Established - Estab
• Expert - Exp
• Trusted - Trust
• Educative - Edu

Using these syllables, you can mix and match to create a unique and meaningful neologism for your brand. Here are a few examples:

• ProInno (Professional + Innovation)
• TechMas (Tech + Mastery)
• QualTrans (Quality + Transformation)
• EduPact (Education + Impact)
• InnoLead (Innovation + Leadership)
• AuthSight (Authoritative + Insightful)
• FutTrai (Future + Training)
• BusAccel (Business + Accelerator)
• ClarExp (Clarity + Expert)
• SucGui (Success + Guidance)

Feel free to mix and match these syllables further to find a name that resonates best with your brand identity and values.

Human Selection

As the LLM has decided, without being prompted to, to give us some examples, we can eliminate and adjust some elements. Here are a few:

• Accel: changed to Ax based on sound similarity,
• Auth: too vague as a single syllable.
• Bus: Biz is more commonly used.
• Con: too vague as a single syllable.
• Det: too vague as a single syllable.
• Edu: too strong a signal towards education.
• Estab: won’t fit well in a brand name.
• Fut: does not sound well.
• Gui: too vague as a single syllable.
• Key: another meaning by itself.
• Mas: too vague as a single syllable.
• Pact: another meaning by itself.
• Pow: too strong signal towards energy.
• Suc: too vague as a single syllable.
• Train: too vague as a single syllable.

So, we are left with: Ax, Ai, Biz, Clar, Exp, Inno, Int, Lead, Pro, Qual, Sight, Tech, Trans, Tru, Val.

With fifteen syllables, picking up two or three for our brand names, we have: 2,940 possibilities. We could ask the AI to try all combination and review them. It is probably more efficient to let the model tray 30 or 50 combinations at a time, hoping that one or two will relate to our brand elements.

Last step

Now, we are ready for the last step of our new brand name exploration. In our next article we will use these syllables, apply the technique that failed last time, hoping that, this time, it will lead to good results. We have no guarantee that we’ll succeed.

When we design AI powered workflows we often end up with a human in the loop. When a workflow is complex enough, require multi-disciplinary, multi-layer thinking, a human decision is necessary. Even in these scenario, using generative AI is a big time saver, leads to structured, well-formed output.

Register to our newsletter to get the next article in your mailbox.

Technology

The above answers were generated by gpt4-turbo, running on azure infrastructure, using our internal web application.

We are Here to Help

At System in Motion, we are committed to building long-term solutions and solid foundations for your Information System. We can help you optimize your Information System, generating value for your business. Contact us for any inquiry.