Why Naming Is Load-Bearing

# Why Naming Is Load-Bearing Let me tell you about Marcus. Marcus spent eight months building a project management tool for freelancers. He called it Trellance — a portmanteau he liked, available on .com, and easy to spell. He built the whole thing. Launched quietly. Started getting traction. Six weeks after launch, a letter arrived. Turns out "Trellance" was close enough to an existing registered mark in Class 42 (software services) that a trademark attorney decided it was worth a cease-and-desist. The company sending the letter wasn't even a competitor. It was a consulting firm in Ohio that had registered the name for internal tooling and never used it publicly. Marcus had three choices: fight it (expensive), rebrand (painful), or shut down (unthinkable after eight months of work). He rebranded. New name, new domain, new social handles, new logo, new email addresses, email migration for existing users, updated App Store and Play Store listings, 301 redirects, and an email to his list explaining why everything looked different. Six weeks of lost momentum on a product that had finally started to move. All of this was preventable. A 90-minute trademark search before he registered the name would have caught the conflict. He just didn't know to look. ## Names are infrastructure Most people treat naming like a decoration. You build the thing, then you pick a name, like choosing what color to paint a room. But the name is actually foundation — it's built into your domain, your social handles, your legal entity, your customer support email, your app store listings, your pitch deck, your business cards, your users' muscle memory. When you change it, you change all of it. And the longer you wait, the more expensive the change becomes. This is what I mean by "load-bearing." The name carries weight. Structural weight. ## The four ways a bad name will hurt you **Trademark conflict.** You can operate for months before anyone notices. Then a lawyer notices. At that point you've already invested everything in the name — financially, emotionally, with users. **Domain unavailability.** The .com is taken. You go with .io or a hyphenated version or a creative spelling. Now every person you tell your name to types the wrong URL. You spend years correcting people. The domain owner knows you need it and prices it accordingly — often $10,000 to $50,000 for a desirable .com. **Handle fragmentation.** You have @yourproduct on Instagram and @yourproductapp on X and @yourproduct_hq on LinkedIn. Your brand is scattered across every platform. Consistency in handles is one of the lowest-cost things you can establish — but only if you check before you commit to a name. **Unsearchability.** You name your product "Arc," or "Notion," or something generic that competes with a thousand other search results. Your SEO is fighting an uphill battle from day one. A distinctive name earns its own search real estate. A generic name makes you invisible. ## Why AI changes this Before AI tools were good, the naming process was either expensive (hire a naming agency) or chaotic (brainstorm with friends until something stuck). Neither was systematic. Neither was fast enough to let you evaluate 50 options before falling in love with one. Now you can generate 50+ name candidates in under an hour, screen them through multiple dimensions simultaneously, and arrive at a shortlist with actual data behind it — trademark status, domain availability, handle parity — before you've registered anything. The workflow this course teaches is systematic. You won't skip the trademark search because it feels tedious. You won't pick the first name that sounds good. You'll have a framework that produces defensible, distinctive names that you've actually checked. ## What AI is not going to do for you I want to be direct about this. AI can generate candidates. It cannot guarantee they're safe. Trademark law is genuinely complex — conflicts depend on class, geography, similarity of goods and services, and prior use in commerce. The USPTO TESS search we'll do in Lesson 4 is a first-pass screen, not a legal opinion. For anything significant (a product you're going to raise money on, a business you're registering), you want a trademark attorney reviewing your final choice. What AI + a systematic process does is eliminate the clearly bad choices before you spend money, and get you to a shortlist that's already much cleaner than what most founders show up with when they finally do talk to a lawyer. ## What this course covers Here's what we're going to do, in order: **Lesson 2 — The Naming Frame.** Before you generate a single candidate, you need a framework for evaluating them. We'll build that framework together. **Lesson 3 — Generating Candidates with AI.** You'll generate 50+ candidates using Claude or ChatGPT directly. I'll show you the specific prompt patterns that produce useful results versus generic ones — and I'll demo what an AI-native naming tool produces so you have a reference for what good output looks like. **Lesson 4 — Trademark Research Without a Lawyer.** The USPTO TESS workflow. Class IDs. What "live" versus "dead" means. Where WIPO comes in for international coverage. **Lesson 5 — Domains and the Extension Question.** .com vs .ai vs .io vs .co. The exact-match myth. What to do when the .com is taken. Domain marketplaces and what's a fair price. **Lesson 6 — Social Handles: The Real Test.** Instagram, X, TikTok, YouTube, LinkedIn, Threads, Bluesky. This is where most names die, and it's often the last thing people check. **Lesson 7 — Narrowing to Three.** You'll have data by now. This lesson is about using it to make a real decision. **Lesson 8 — Brand Identity with AI.** Logo directions, color palettes, typography. Using AI to generate visual identity options without hiring a designer for the first pass. **Lesson 9 — The Brand Brief.** A 1-page document that captures who you are and why you chose this name. The thing that makes every future branding decision faster and more consistent. **Lesson 10 — Naming the Future.** The principles that survive across products and pivots. When to rename. What's next. ## What you'll need - Something you're naming. It doesn't have to be real yet — a half-baked idea counts, a side project counts, a thing you've been thinking about for a year counts. You just need something to run through the process. - A Claude or ChatGPT account (free tier is fine). That's the tool you'll use to generate candidates in Lesson 3. Lessons 3 and 8 also include an optional Swishh demo embed so you can see what an AI-native brand-builder produces — but you don't need a Swishh account to complete either lesson. - Access to USPTO TESS at tess.uspto.gov — free, no account required. - About an hour per lesson. Some lessons are shorter, some longer. The exercises take real time. That's the setup. Let's build something worth naming.