Naming Your Business: A Founder’s Practical Guide

Choose your business name and domain together. Generate several usable options, eliminate obvious conflicts, test whether people can say and spell the finalists, and get trademark clearance before committing substantial money or launching. For a new U.S. startup, an affordable matching .com is a strong starting constraint.

Here is a process you can use with your cofounder. The target is a short list of workable names, a documented choice, and control of the domain and brand assets.

What should a good business name do?

Seth Godin’s “Naming a business” makes a stronger recommendation than simply avoiding a literal description: choose a name disconnected from your category. Customers will attach meaning to it through their experience of the company. He favors real words, easy spelling and pronunciation, and a tagline that explains the business.

Our recommendation is to keep the search broad. Include familiar words with no obvious connection to the business, as well as words or images that could support the brand or suggest a customer outcome. A connection can be useful, but it is not a requirement. The business can give the name its meaning over time.

There is a legal reason to consider distinctive names. The USPTO identifies suggestive, arbitrary, and invented marks as stronger than descriptive or generic terms. Invented words can be protectable even though Godin prefers real ones. The challenge is making them easy to use.

Any candidate still has to work in ordinary situations: a customer recommends you aloud, someone types your email address, or a buyer tries to distinguish you from another company.

Why does Paul Graham push so hard for the .com?

In “Change Your Name”, Graham argues that U.S. startups should generally rename if they do not own the matching .com. He sees the domain as a credibility signal and founders’ resistance as a combination of attachment and difficulty imagining alternatives.

The strongest reading of his argument is practical. Before launch, a founder may be deeply attached to a name that has little accumulated customer recognition. Other names could serve the company just as well. Searching more broadly may produce a good name with an obtainable .com and avoid years of explaining a workaround.

That does not require paying any price for your current name. It requires considering other names.

A public reproduction of Graham’s early-October 2026 post captures the idea:

“The .com” is just the only .com you’ve been able to think of so far.

His recent comments about finding a better .com for $5,000 are a challenge to widen the search, not a guaranteed price or a spending target.

Our recommendation is to start with names whose matching .coms are obtainable within your budget. Choosing another extension can be reasonable, particularly for an established brand or a specific audience. Make that choice after comparing viable alternatives and checking who controls the matching .com.

How do you generate a useful shortlist?

Start with a brief that fits in five lines:

• Customer: Who needs to recognize and remember the name?
• Scope: What do you sell today, and what might you sell next? Use this to screen for constraints, not to supply naming keywords.
• Markets: Which languages and countries matter at launch?
• Budget: What can you spend on the domain, separately from legal review and branding?
• Decision owner: Who chooses if the founders disagree?

Set the budget before seeing asking prices. A domain purchase should fit the operating plan.

Then schedule one working session to generate roughly 30 candidates. That is a suggested exercise, not a research-backed minimum. Draw from three sources:

1. Familiar words and unexpected pairs. Browse objects, foods, animals, plants, colors, and places, then try combinations. Include options with no product connection. CopperKite, the fictional candidate below, illustrates this approach.
2. Brand imagery and customer outcomes. Consider words or images that could support the brand you want to build, or suggest an experience such as confidence, ease, or momentum. Use these as another source of candidates, without requiring every name to communicate a benefit.
3. Books, pictures, and surroundings. Look through a novel, a photo collection, or the objects around you. Collect words and images that catch your attention, including ones outside your business category.

Keep candidates you can imagine saying naturally in an introduction. Compare both unrelated and suggestive names; do not require a product connection or reject a usable name merely because it has one.

Write the category explanation separately: “[Name] helps [customer] do [job].” Let the tagline and website explain the offering while the name identifies the company.

AI can generate candidates. Verify its claims about domains and trademarks independently.

YC’s broader advice emphasizes launching and talking to users. Apply that discipline here: book a second session to choose finalists, rather than keeping an open-ended naming discussion alive for weeks. Legal review may need its own timetable.

How should you compare the candidates?

Use one row per name. Record observations rather than a single numerical score that can hide a serious problem.

Check	What to record
Say it	How people pronounce the written name
Spell it	How people spell the name after hearing it
Remember it	Whether they recall it after a short unrelated conversation
Grow with it	Whether it ties you to a feature, technology, or geography you may outgrow
Distinguish it	Similar businesses, names, and search results
Obtain the domain	Matching .com, current use, asking price, and transfer feasibility
Assess legal risk	Initial search findings, unresolved questions, and counsel’s clearance assessment

Test a few finalists with five to ten people who resemble your customers. This is an informal usability check, not a statistically validated study.

Say the name without showing it and ask them to write it. Separately, show it and ask them to pronounce it. Later, ask what they remember. Avoid explaining the clever spelling before the test or asking only whether they like it.

Repeated mistakes are reasons to reconsider. A name that requires coaching will require coaching outside the test too. For important foreign-language markets, include speakers from those markets.

Which problems should eliminate a name?

Drop or pause candidates when:

• People repeatedly mishear, misspell, or confuse them with another business.
• The domain exceeds your budget or depends on an owner who has not agreed to sell.
• The name creates a misleading association or ties you to a feature, technology, or geography you may outgrow. Being unrelated to the product is not a reason to reject it.
• Initial searches reveal a serious potential trademark conflict that remains unresolved.

An attractive logo or enthusiastic founder vote should not override those problems. Where a legal finding is ambiguous, have counsel assess it before treating the candidate as available.

What does this look like in practice?

Imagine a startup selling invoice software to small agencies, with plans to add payments and cash-flow tools. Its domain budget is $2,500.

The names, prices, and test results below are invented to illustrate the decision. No availability or trademark research has been performed on these candidates.

Candidate	Assumed findings	Decision
InvoicePilot	Easy to understand; matching .com quoted at $18,000; name centers on the first product	Reject under the stated budget; keep generating alternatives
Klyrr	Matching .com quoted at $600; several testers spell it “clear” or “clir”	Reject because repeated spelling confusion undermines the price advantage
CopperKite	Testers say and spell it consistently; matching .com quoted at $1,800; no obvious conflict in the initial screen	Advance to legal review; broad enough to support the planned products

CopperKite is the strongest finalist under these assumptions. Its lack of connection to invoicing leaves room for the business to evolve. “Invoicing and payments for independent agencies” explains the offering. There is no need to invent a story connecting copper or kites to financial software.

The next action is legal review, followed by acquisition if the findings and transaction terms are acceptable. If a serious conflict emerges, the founders return to their list. They do not change one letter and assume the problem has disappeared.

What checks are needed before committing?

There are four separate questions:

• Entity name: Will the formation state accept the legal company name?
• Domain: Can the company register or acquire the address?
• Trademark: Can the company use and seek protection for the brand given existing rights?
• Trade name or DBA: Are filings required if the public brand differs from the legal entity name?

The SBA explains that these registrations serve independent purposes. State acceptance and domain ownership do not clear a trademark.

For an initial screen, search the web, relevant industry directories and app stores, and the USPTO trademark database. Search close spellings, sound-alikes, and related goods or services as well as exact matches.

The USPTO’s confusion analysis considers similarities in sound, appearance, meaning, and commercial impression. Different trademark classes do not automatically resolve a conflict. Its clearance guidance also recommends looking for commercial use outside the federal register.

Send counsel the finalists, proposed goods and services, target markets, domain findings, and potential conflicts you found. Ask for a clearance assessment and filing recommendation before a premium domain purchase, substantial branding work, or public launch. Include foreign jurisdictions where relevant.

This belongs in your incorporation planning, rather than waiting until the company is already using the name everywhere.

When can you stop searching and launch?

Choose once a candidate works in the usability checks, fits the foreseeable business, has an obtainable domain within budget, and has received an acceptable legal assessment. Do not keep searching solely because someone might someday imagine a name everyone likes better.

Before launch:

• Complete the domain registration or purchase, using appropriate payment and transfer safeguards for a material acquisition.
• Put the registrar account, renewals, and recovery credentials under company control, with a backup administrator and multifactor authentication.
• Document any transfer from a founder who acquired the domain personally.
• Address ownership of contractor-created logos and other brand assets in signed agreements. Keep these with your IP ownership records.
• Complete applicable entity-name or DBA filings and implement the trademark filing strategy agreed with counsel.

Keep the clearance materials and acquisition records in the company’s data room. Then use the name consistently and get back to customers.

Frequently asked questions

Can I register a cheap domain while I evaluate a name?

You can register an ordinary available domain as a provisional candidate, but registration gives no trademark clearance. Investigate known conflicts first, and avoid substantial acquisition costs or public brand investment until the candidate has been assessed.

Should I buy several domain extensions?

Consider the extensions or obvious variants customers are likely to use, within a defined budget. Owning a collection of domains does not substitute for trademark clearance or control of your primary address.

Does my legal entity name have to match the brand?

Not necessarily. Check applicable trade-name or DBA requirements, and use the correct legal entity in contracts so customers know who they are dealing with.

---

This post is for general informational purposes only and does not constitute legal advice. Trademark rights, name availability, and filing requirements depend on the jurisdiction and specific facts. Consult qualified counsel about your situation.

Have a shortlist ready? Book a free call to discuss clearance, ownership, and the steps to take before launch.