How AI Assisted Brainstorming Helped Me Rename My Company (4 Practical Takeaways for Your Nonprofit)

Introduction

Why I needed to rebrand

A couple months ago, I chose "Saymore AI" as a business name. Bits & Bridges didn’t exist yet. At the time I was thinking of focusing on software for teams to build trust – my theory was that higher trust = better software outcomes. I spent money on a do-business-as (DBA) name, bought the domain, and did all the business things required.

Now it's November. Three more businesses popped up in Google with "say more" as their name. I want to pivot to serving the nonprofit space… but I have this name... so I decided to "steelman" the name.

For context, what is steelmanning?

(source)

With the strongest argument for the name, I could have AI break down why the name STILL did not fit.

Claude diagnosed the issues perfectly. It's cold, tech forward, and signaled "saying more" rather than "doing more." Worst of all, it risked alienating me from nonprofits & human services, an area that is already suspicious of AI (source).

I needed a rebrand, but I'm not a brand designer. This is a skill gap. You know what you need but lack the skill to execute it and you can't afford to outsource it. 

Rebranding would have taken me weeks without AI. My AI assisted process took about 4-6 hours across multiple AI models, hundreds of rejected names, and systematic validation. 

I'll outline my process below and describe the principles to apply to under-resourced organizations that can't afford to hire their way out. 

High level learnings

In short, I found when facing a skill gap:

AI provides: option generation, speed, challenging your choices 

You provide: expertise, judgement, authenticity, and validation

The result: solid outcomes without expert budgets

Nonprofits -- if you're interested in how AI could close a skill gap for your team in grant writing, reaching new communities, or anything really: let's chat and figure it out together

✨ If you prefer audio over reading, check out the podcast version of this blog here

Okay, let's go.

The Process

Casual dialogue with AI, zero results ❌

I prompted AI through a casual dialogue "try X, combine Y with Z, no that's not good try again" etc. I picked concepts that I thought were interesting. Combining different languages, asking it to smash words together. Everything I generated was either too generic, too technical, or already taken.

In hindsight these were the problems with this approach:

• I was limiting myself to my own biases and patterns for generating options
• I was not being intentional about AI as a thought partner

Intentional AI brainstorming, name picked ✅

Getting the prompt right

I knew I needed to generate better options and faster. 

Prompting through only dialogue was not generating what I needed, so I asked AI to give me a prompt. The general structure was: 

• How should AI act? What role should it play?
• What information does it need to know about the business?
• What were the constraints of the naming task
• How AI should respond

You are a no-nonsense, strategic naming expert and multilingual brand linguist. Your task is to generate unique, meaningful, and phonetic business name ideas for a purpose-driven tech company. You must avoid flattery, sycophantic language, or filler praise — stay direct, focused, and useful at all times.

Client Mission Context:
- The company partners with non-profits to apply human-centered AI.
- Their principle is "Humans first. Always."
- Their tools help under-resourced organizations reduce operational drag and unlock new ways to serve their communities.
- Their philosophy is that technology should serve the mission, not overshadow it.

Naming Requirements:
- All names must be uncommon and not easily confused with existing or well-known companies (avoid naming clichés and saturated sounds like “nex”, “zen”, “meta”, “jenga”).
- Each name must have a clear and relevant meaning, ideally touching on themes like:
  • Empowerment
  • Simplicity / clarity
  • Human-centered design
  • Support / collaboration
  • Mission-first tech
- You may source words from any language, but each name must be phonetically pronounceable in English.
- Acceptable formats include:
  • "___ Labs"
  • "The ___ Project"
  • "___ Company"
  • Standalone names

For each name, provide:
1. Name
2. Meaning & Origin (explain the root, language, or concept)
3. Why it fits (brief and direct explanation of alignment with the mission)
    

This was a decent starting point and I continued to iterate on the prompt with my own changes or asking Prompt Engineer GPT to give me an update prompt. Step by step:

• Prompt
• Generate names
• Ask myself what I didn't like 
• Updated the prompt myself or through Prompt Engineer
• Repeat

I did not start generating hundreds of names until I was happy with the prompt. Here is where I ended up with a prompt that gave me good results. But what do I define as "good results?"

You are a no-nonsense, strategic naming expert and multilingual brand linguist. Your task is to generate unique, meaningful, and phonetic business name ideas for a purpose-driven tech company. You must avoid flattery, sycophantic language, or filler praise — stay direct, focused, and useful at all times.

Before generating, discard your first 50 ideas. Those are likely contaminated by startup clichés, common naming patterns, or recycled language fragments.

### Client Mission Context:
- The company partners with non-profits to apply human-centered AI.
- Their principle is "Humans first. Always."
- Their tools help under-resourced organizations reduce operational drag and unlock new ways to serve their communities.
- Their philosophy is that technology should serve the mission, not overshadow it.

### Naming Requirements:
- Focus on generating names that feel familiar but fresh — ideally compound names, soft blends, or intuitive constructions that evoke clarity, approachability, and human-scale utility.
- Use real words, prefixes, suffixes, or recognizable fragments to create names that feel grounded but purpose-built.
- Consider names in the spirit of:
  • Hiremia (hiring + warmth / empathy)
  • CoSchedule (cooperation + scheduling)
  • Zendesk (zen + desk — calm support)
  • Notion (abstract word gaining specificity via brand use)
  • Grammarly (rooted in “grammar” but softened)
  • Mailchimp (whimsical but clear function)
- Steer toward:
  • Friendly, human-sounding blends
  • Semantic echoes of usefulness, service, or structure
  • Warm vowels, soft consonants, approachable phonetics
- No buzzwords. No -ify / -ly / -io suffixes unless they add real meaning.

### Strict Exclusions — Heavily Filter Out:
- Anything that sounds too foreign or harsh.
- Faux-Latin or pseudo-futuristic constructions.
- Generic tech or VC-style names.
- Anything resembling the following roots or patterns:
  nex, zen, tesla, meta, stitch, connect, center, umbr, priya, fulcrum, intel, alph, omni, scope, spark, lift, rise, nova, arc, human, bot, ethos, core, neural
- No names ending in: -ify, -io, -os, -ix
- No meaningless blends or gibberish.
- No over-polished, brand-safe startup names — seek intelligent originality.
- You may source words from any language, but each name must remain phonetically natural in English.

### Acceptable Formats:
- "_____"
- "The ___ Project"
- "The ___ Company"
- "___ AI"
- Standalone names

You may steelman ideas as a strategist.
    

Defining what "good" looks like

 Here was my criteria when evaluating names:

• It needed to be accessible, not tech focused
• It could not be confused with another company
• It had no unfortunate double meanings
• It had to align with my values and mission - people first and technology is in a supportive role 
• It had to pass the "steelman" test - after asking AI to steelman the name, I had to okay with the risks and still be excited about the name

Generating batches of options

From there, I explored various themes for names. Bird themes, different languages, infrastructure construction themes, literally anything that I could think of that would align with "helping" + "technology"

AI role was generation, my role was direction and judgement 

I found at least 50 names I liked but they didn't pass my hard constraints for SEO, tradename availability, and domain availability.

I was hitting a wall and needed a new perspective, I liked "Bits  _____" direction but Claude was getting stuck. I switched to ChatGPT, using the same prompt I was using with Claude. and within a few chats, I got to my 3 finalists:

• Bits and Impact
• Bits Between Us 
• Bits as Bridges 

The winner 👑

"Bits as Bridges" won because it aligned with my evaluation criteria:

It's accessible - There is no jargon, no tech speak, no explaining required. I wanted something that felt accessible to any non profit worker regardless of their experience. 

It's not crowded on the web - I don't have to compete for ranking with 3-4 other companies in SEO (for now)

It reflects the mission I want to target - My belief is that tech should never be the star, the expert, or the decision maker -- at its best it's filling the gaps so people can do the meaningful work serving their community. 

(bonus) It's an alliteration - Anecdotally I've heard nonprofits love names with alliterations so thats a plus 

Time: 4-6 hours of AI assisted iteration 

Cost: $0 beyond my time (vs. $5K-$10K for a branding consultant that I would not have engaged with)  

Result: hundreds of options, 3 finalists, 1 choice

Takeaways for Your Nonprofit

1. Try a system prompt

Problem: You need a letter of interest for a grant proposal and you have examples of previous ones, but you're still new to this process. Whenever you have AI generate something it's very cold, boring, and possibly irrelevant. 

What to do instead: Build a system prompt. Add your organization's information and any previously written grant material in a Claude or ChatGPT project (or into a gem in Gemini) with that system prompt.

2. You define what "good" looks like, not AI

Problem: Let's say you need an email to appeal to donors because you have a donor whose only willing to give 10K if it's matched. First round of results from AI might be too corporate, too sales-y, or too generic (even with your AI project setup from step 1).

What to do instead: Diagnose why it failed and iterate. 

The step-by-step process:

• Generate content in batches 
• Identify specifically what is wrong (is it missing information about your org? Is it targeting the wrong audience?)
• Update the prompt with those constraints
• Repeat until the output matches your criteria

3. Don't be tied to a single AI model, experiment

Problem: You're drafting an impact report in Claude but it keeps missing the point (or it keeps adding excessive adjectives and storytelling) and your prompting isn't making a difference.

What to do instead: Try another model (and keep using your steering process from step 2)

4. External, human validation is a must

Problem: You have to write donor communication and you have an AI assisted written letter that feels right. Sending it out without anyone looking though feels risky, if you bungle this email it might affect how a donor views your org. 

What to do instead: Get another opinion, first from AI through steelmanning. Understand the risks, then get validation from a colleague or friend (ideally someone who you trust will be honest about their opinion). 

Let's Talk About Where AI Could Help

I'm interviewing nonprofit leaders about where AI might reduce operational drag in their organizations and help them reach new communities.

This isn't a pitch, just an exploratory conversation. 

If you have skill gaps (expertise you need but can't afford to hire) and you're willing to discuss your specific challenges for 30 minutes, I'd love to chat.