How to Name Your Brand: From Brainstorm to Trademark (2026)

In July 2023, Elon Musk deleted one of the most recognized brand names on the planet. Twitter became X overnight. Brand Finance tracked what happened next: the name's value dropped from $5.7 billion to $673 million in under two years. An 88% collapse. Two years on, more than half of Americans still call it Twitter.

Musk could absorb that loss. You can't.

Your business name is the most permanent branding decision you'll make. Change your logo next quarter. Rebuild your website over a weekend. But try changing your name after customers know you, and you're essentially starting over — with the added cost of explaining why.

And it doesn't have to be a billion-dollar company to matter. A coffee shop owner in Astoria, Oregon named her shop "Sambucks" after her maiden name, Sam Buck. Starbucks sued her and won. Even your own name can become someone else's trademark problem.

This article gives you the full naming pipeline: how to brainstorm names (with and without AI), how to evaluate them, what trademark protection actually costs, and what to do when every domain you want is already taken.

What's in This Article

• Why Your Business Name Is Worth Getting Right
• The Five Types of Business Names
• The Naming Pipeline: From Brainstorm to Trademark
• Using AI for Brand Naming
• When the .com Is Already Taken
• Evaluate Your Brand Name
• Frequently Asked Questions

Why Your Business Name Is Worth Getting Right

When Jeff Bezos incorporated his first company in 1994, he called it Cadabra, a play on "abracadabra." His lawyer, Todd Tarbert, pointed out that on the phone, it sounded like "cadaver." Bezos went to the dictionary, landed on "Amazon" because it was the world's largest river and he wanted to build the world's largest bookstore, and registered the domain.

Phil Knight wanted to call his shoe company "Dimension Six." His first employee, Jeff Johnson, dreamed up the name Nike, after the Greek goddess of victory, the night before their shoe boxes were due to print. Knight wasn't thrilled. He went with it anyway. Nike's revenue hit $51.4 billion in 2024.

Those are fun stories, but they obscure something important: for every name that worked, there's a business that got it wrong and paid real money to fix it.

Monster Energy sued a small aquarium products company called Monster Fish Keepers. Different industry, completely unrelated products. The settlement cost the small business $65,000 in legal fees and a complete rebrand. CarMax faced a trademark suit from TJ Maxx over the word "Max." Even though CarMax won, the legal fees ran to about $1 million.

Marty Neumeier, author of The Brand Gap, puts it directly: "The wrong name can cost millions in workarounds and lost income over the lifetime of the brand."

And David Placek, founder of Lexicon Branding (they named BlackBerry, Dasani, and Swiffer), says something that should comfort anyone agonizing over this decision: "Great names make you uncomfortable. Sonos was rejected, Azure was called 'dumb,' BlackBerry was deemed 'crazy.'"

So if you hate every name on your shortlist, you might actually be onto something.

The Five Types of Business Names

Not all names are created equal, legally or strategically. Understanding the type you're choosing helps you avoid the two most expensive mistakes: picking a name you can't trademark, or picking one that doesn't travel.

Invented names (Google, Xerox, Kodak) are the strongest from a trademark perspective. No one else has the word, so it's yours to defend. The downside is they mean nothing until you spend money making them mean something. Google was a misspelling of "googol" that happened when a grad student checked domain availability. It worked because the product worked. The name followed.

Real-word names (Amazon, Apple, Slack) borrow meaning from an existing word and redirect it. When it works, it's powerful. Apple suggests simplicity and approachability in a way that "Integrated Business Machines" never could. But the more common the word, the harder it is to trademark in your industry, and the more likely the .com is already gone.

Descriptive names (General Electric, PayPal) tell you what the company does. Business owners love these because they feel safe. But they're a trap. The USPTO's most common reason for rejecting a trademark application is that the name is "merely descriptive." You can't own a generic phrase. If you call your plumbing company "Quality Plumbing," you'll never be able to stop someone else from using a similar name.

Founder names (Ford, Disney, Goldman Sachs) carry personal stakes. They work when the founder IS the brand. For most small businesses, they limit scalability and make the company harder to sell. Honestly, I'm biased on this one. If you name your company after yourself and you're not already well-known, you're betting that your personal name alone will mean something to strangers. Most of the time, it won't.

Acronyms (IBM, BMW, UPS) are the weakest starting position. They mean nothing, sound like nothing, and are nearly impossible to trademark unless they've already achieved recognition. IBM works because IBM has been around since 1911. Your new consulting firm's three-letter abbreviation won't have that advantage.

Here's the trade-off at a glance:

Name Type	Trademark Strength	Domain Availability	Instant Meaning
Invented (Google, Xerox)	Strongest	High	None
Real Word (Amazon, Apple)	Strong (in unrelated category)	Low	Metaphorical
Descriptive (PayPal, General Electric)	Weakest	Very Low	Immediate
Founder (Ford, Disney)	Moderate	Moderate	Personal
Acronym (IBM, BMW)	Weak until famous	High	None

If you're starting a business and want the strongest possible foundation, invented names or real-word names in an unrelated category give you the most room to grow and protect.

The Naming Pipeline: From Brainstorm to Trademark

Every naming article tells you to "brainstorm." Almost none tell you what happens after that. Here's the actual process, with what each step costs.

Step 1: Generate a lot of names

David Placek's team at Lexicon generates 1,000 to 1,500 names before finding gems. You don't need that many, but you need more than five. Aim for 50-100 starting candidates.

Use these prompts to push beyond the obvious:

• What metaphor describes what your business does for people?
• What words from outside your industry capture the feeling you want customers to have?
• What would your business be called if it were a restaurant? A band? A planet?
• What two unrelated words could you combine? (Instagram = "instant" + "telegram." Netflix = "internet" + "flicks.")

The goal isn't to find THE name right now. It's to find 8-10 contenders worth evaluating. (AI tools can help here — I'll cover what works and what doesn't in the next section.)

Step 2: Kill the weak ones fast

Neumeier's checklist is the best shorthand I've found. A strong name should be: short (four syllables or fewer), easy to spell, satisfying to say out loud, and different from your competitors. If 90% of businesses in your industry use similar-sounding names (and according to naming expert Steve Manning, they do), yours should sound nothing like them.

Run each surviving name through these filters:

• Does it have an embarrassing meaning in another language?
• Can someone easily misspell it?
• Will a customer remember it after hearing it once?

Picture this: a retail store owner walks in with a shortlist of ten names, and eight of them are variations of "[Location] + [Product Type]." Austin Home Goods. Hill Country Living. Capital Furnishings. They're descriptive, forgettable, and impossible to trademark. The two names that don't follow that pattern? Those are the ones worth testing.

Step 3: Check domain availability

This is where most naming processes stall. There are over 159 million .com domains registered. Only about 1.6% of English dictionary words are still available as .com addresses. If your dream name's .com is taken, don't panic. (I'll cover what to do about that in the next section.)

Step 4: Run a trademark search

This is not optional.

Almost half of all USPTO trademark applications get rejected. The success rate has dropped from 59% to about 52% over the last five years, partly because there are now over 3.3 million active registrations competing for space. The most common rejection reason? "Merely descriptive."

A basic search through the USPTO's free TESS database catches obvious conflicts. But a comprehensive search by a trademark attorney costs $300-$600 and catches the non-obvious ones. That's the search that would have saved Monster Fish Keepers $65,000.

I used to think the DIY trademark route was fine for most small businesses. I've changed my mind on that. Self-filed applications succeed only about 46% of the time, compared to roughly 60% for attorney-filed applications. That gap alone can justify the legal fee.

Step 5: File the trademark

As of January 2025, the base USPTO filing fee is $350 per class of goods or services. Add a trademark attorney ($750-$2,500) and a comprehensive search ($300-$600), and you're looking at $1,050 to $3,500 all in for a single-class trademark.

That might sound like a lot. It isn't. It's insurance. The alternative is spending years building brand equity under a name someone else owns, then getting a cease and desist letter that costs $10,000-$20,000 just to respond to.

Here's what the investment looks like at each level:

Approach	Cost	What You Get	Risk Level
DIY + free tools	$350-$500	Your own brainstorm, TESS search, USPTO filing	High (no professional search)
Crowdsource contest	$300-$750	100+ name ideas, audience testing available	Moderate (creative input but no legal vetting)
Boutique branding agency	$5,000-$15,000	Naming + trademark search + brand strategy	Low (professional process, legal clearance)
Specialist naming firm	$50,000-$250,000+	Full research, linguistic testing, cultural analysis	Lowest (enterprise-grade process)

For most business owners reading this, the DIY path works if you're disciplined about the trademark search. But if your business is generating real revenue and the name is going on everything (signage, vehicles, packaging, a building), the boutique agency tier buys you a name that's been professionally cleared and strategically developed. That's not a luxury. It's the difference between building on solid ground and building on a name that might not be yours to keep.

If you want to understand what goes into a full brand identity beyond just the name, our brand strategy guide covers the bigger picture.

Using AI for Brand Naming

Let's address the obvious question. You've probably already typed "business name ideas for [your industry]" into ChatGPT. Most business owners do. Here's what you need to know before you trust what comes back.

Where AI actually helps

Brainstorming volume. AI is genuinely good at generating large quantities of name candidates fast. NameStormers, a naming agency with 40+ years of experience, tested ChatGPT for naming waterproof socks and got functional options like "DryStride," "AquaStep," and "StormSox" within seconds. When pushed for creativity, it produced abstract options like "Aquilo" and "Nimbus Knit." If Lexicon's team generates 1,000-1,500 names before finding gems, AI can get you to 200 candidates in an afternoon instead of a week.

Breaking creative ruts. When you're stuck on variations of "[City Name] + [Industry]," AI can pull from unexpected directions. Ask it for metaphors, mythology references, or sounds that evoke a specific emotion. The lateral thinking is where it earns its keep.

Preliminary screening. AI can flag obvious linguistic problems — a name that sounds unfortunate in Spanish, or one that's uncomfortably close to a well-known brand. Not comprehensive enough for a final check, but it catches the obvious misses early.

Where AI falls apart

The "Name Slop" problem. NameStormers coined this term for a real pattern: AI-generated names that look plausible but don't actually mean anything. No story behind them, no emotional hook, and often no legal safety. Because AI draws from the same training data, it gravitates toward the same patterns — that's why suddenly every other startup name ends in "-ly," "-io," or "-fy."

Everyone gets the same suggestions. This is the part most people miss. If you and three competitors each ask ChatGPT for "modern, memorable names for a sustainable clothing brand," you'll get overlapping lists. AI doesn't know who else is asking the same question. When I see naming shortlists from different business owners in the same space, the overlap is genuinely startling.

Trademark blindness. This is the dealbreaker. AI does not check trademark databases. And the problem goes deeper than just not checking — AI generates names by recombining patterns from its training data, which includes existing brand names. The National Law Review calls this "algorithmic genericism": names that sound like brands because they echo ones that already exist. "Glowsense" for skincare. "Verdana" for a wellness brand. Names that feel right precisely because they're derivative.

And unintentional infringement is still infringement. Under trademark law, it doesn't matter whether an AI generated the name or you dreamed it up yourself. If consumers might be confused, liability exists.

No real consumer testing. Mike Carr at NameStormers, who's spent four decades in naming, puts it directly: "You cannot test the name with synthetic respondents." AI can tell you what a name sounds like in theory. It cannot tell you how real humans in your target market will react to it.

Which tools to use (and how)

ChatGPT and Claude are the best starting points for creative brainstorming. They're conversational and you can iterate. The key is giving them serious context — not "give me business names for a bakery" but a detailed brief: your target customer, your competitive landscape, your price positioning, and what feelings you want the name to evoke. One prompt won't do it. Plan on spending an hour across multiple rounds.

Dedicated naming tools like Namelix, Shopify's Name Generator, and Looka check domain availability and can generate logo mockups alongside names. Faster for volume, but their output tends to be more formulaic because they optimize for "looks like a brand name" rather than "is the right name for your specific business."

Squadhelp (now Atom.com) runs a hybrid model — AI generation plus human crowdsourced contests plus a curated domain marketplace. The creative range is wider than pure AI, but quality varies wildly. One user who bought the premium package received 2,100 name submissions and described only six as "not terrible."

Here's an honest comparison:

Tool	Best For	Watch Out For	Cost
ChatGPT / Claude	Creative brainstorming, exploring metaphors, iterating on concepts	No trademark or domain checks; quality depends entirely on your prompts	Free – $20/mo
Namelix	Quick name + domain check + visual mockups	Formulaic output, same suggestions given to many users	Free
Shopify Name Generator	E-commerce store names with instant domain availability	Basic output, built primarily for Shopify merchants	Free
Looka	Name + logo + brand kit in one workflow	Name suggestions tend toward generic; logo quality varies	Free for names; logos from $20
Squadhelp / Atom	AI + human hybrid, premium domain marketplace	Contest quality varies wildly; premium domains can run to five figures	$300–$750 for contests

The smart approach

Use AI for Step 1 of the pipeline (generating candidates) and nothing else. Feed it a detailed brief. Generate 100-200 names. Then apply the human filters from Steps 2 through 5 yourself — or with a professional.

Here's a prompt structure that actually produces useful output:

"I'm naming a [type of business] that serves [target customer]. Our competitors are [list 3-5]. Our price positioning is [budget/mid/premium]. The feeling I want customers to associate with the name is [2-3 adjectives]. Generate 30 name candidates across these categories: invented words, real words used metaphorically, and unexpected combinations. Avoid names ending in -ly, -io, or -fy."

Run that three or four times with slight variations, and you'll have a solid candidate pool to start filtering.

The one thing AI absolutely cannot do is tell you whether a name is legally safe to use. And that's the one thing that actually matters.

When the .com Is Already Taken

The good news: your perfect name doesn't need a .com to succeed.

The bad news: most business owners still think it does.

Here's where it actually matters. If your customers skew older or you're in a conservative industry (law, accounting, healthcare), the .com extension carries measurable trust. 76% of consumers say familiarity with a domain extension increases their trust. For a local service business targeting that demographic, that matters.

If you're targeting younger customers or running an online-first business, the trust penalty is shrinking every year. Companies like Intercom and Redis built billion-dollar businesses on .io domains.

WIPO processed a record 6,200 domain name disputes in 2025, the highest caseload in its history. Domain squatting is growing. If your business name is already registered by someone sitting on it, you have three options:

Modify the domain, not the name. Add "get," "try," "use," or "hello" before your name. Slack launched on getslack.com before acquiring slack.com.

Use a modern TLD. The .co, .io, and .shop extensions are all viable. Just make sure you also register the .com if possible (even if it only redirects) to prevent someone else from capitalizing on your brand.

Buy it. If the .com is parked and not in active use, you can make an offer through a registrar's transfer service. Expect $500 for a forgotten domain up to five or six figures for anything short and memorable.

To be fair, there's no universal right answer here. A local boutique doesn't need a three-letter .com. A SaaS company launching nationally probably should invest in one.

Evaluate Your Brand Name

Before you commit to a name, score it. This scorecard distills Neumeier's criteria, trademark considerations, and domain realities into a single evaluation you can run on any name in your shortlist. Rate each factor from 0 (weak) to 2 (strong).

Not Sure Where to Start?

Naming a business sounds like a creative problem, but it's really a strategic one. The creative part is the fun half-hour of brainstorming. The strategic part (trademark clearance, domain strategy, making sure the name supports where your business is going in five years) is where the real work lives.

If you're stuck in the brainstorm loop or you want someone to evaluate your shortlist, book a free 15-minute call and we'll tell you what we'd do. Or if you want to understand what the full naming-to-identity process costs, see our breakdown of logo design pricing.

Frequently Asked Questions

How long does the naming process typically take?

For a DIY approach, expect 2-6 weeks from first brainstorm to trademark filing. With an agency, the timeline is usually 4-8 weeks, because it includes strategy sessions, creative rounds, and legal clearance. The USPTO itself takes 6-8 months to process a trademark application after you file, but you don't need to wait for approval to start using the name. You have common-law rights from the moment you begin using it in commerce.

Can I trademark a name if someone else already uses it in a different industry?

Sometimes. Trademark protection is industry-specific, so "Delta" can exist as both an airline and a faucet company. But there are limits. If the existing mark is famous enough (think Apple, Nike, or Monster Energy), courts may block your use even in a completely unrelated category. A trademark attorney can evaluate your specific situation for $200-$500, which is significantly cheaper than learning the answer through a lawsuit.

Should I test my name with customers before committing?

Yes, but keep it simple. Show 3-5 name options to 15-20 people in your target market and ask two questions: what does this name make you think of, and would you trust a company with this name? Don't explain the names first. If they need explaining, they're not working.

Can I just use ChatGPT to name my business?

You can use it to brainstorm — and you should. AI is excellent at generating volume and pushing past your first obvious ideas. But don't stop there. AI doesn't check trademarks, doesn't know what names it's suggesting to your competitors, and tends to generate similar-sounding options across users (the "-ly" and "-io" epidemic). Use AI for the first 20% of the process (generating candidates), then apply human judgment for the other 80% (evaluating, testing, legal clearance). The biggest risk isn't that AI gives you a bad name — it's that it gives you a name someone else already owns.

What if I already have a name but I'm worried about trademark issues?

Start with a free search on the USPTO's TESS database (tess.uspto.gov). If you find similar marks in your industry, consult a trademark attorney before investing further in the name. The earlier you catch a conflict, the cheaper it is to resolve. Rebranding a five-year-old business is exponentially more expensive than renaming one that hasn't launched yet.