NameRobo
AI Domain Name Ideas Generator
Why generating ideas is harder than checking them
Most founders exhaust their candidate list before they exhaust the availability checker. Brainstorming and naming are different skills: brainstorming generates volume without filters; naming applies constraints — character limits, memorability, trademark viability, and available TLDs — that shrink the pool dramatically. The gap between "sounds good in a room" and "passes the full checklist" is larger than most people expect. Tools that only check availability solve half the problem; the harder half is generating candidates worth checking in the first place, which requires deliberate frameworks rather than free association.
Frameworks for generating candidates worth checking
Root-and-suffix exploration starts with one strong word and systematically pairs it with prefixes and suffixes to build a shortlist: "arc", "arcly", "arco", "arcform", "arcpath". Semantic neighbor mapping lists synonyms and related concepts for your core value proposition, then tests each for domain availability. Concatenation combines two short unrelated words that create an interesting collision ("taskbird", "clearform", "driftlog"). Foreign-language root borrowing lifts phonetically appealing words from other languages — Italian, Portuguese, and Finnish are popular sources — but requires a trademark search since the same word may already be claimed in another market.
From shortlist to registered domain
Stop accumulating candidates when you have 15–20 names that pass a quick gut filter. Score them against four axes: pronounceability by someone who has never seen it, spellability after hearing it spoken once, trademark search results, and the availability of matching social handles. Run the top five through a "partner test" — say each name to someone outside your industry and ask what business they think it is. The winner is the name that scores highest across all axes, not the one that sounds best to you alone. Register immediately; the window between decision and purchase is where squatting risk peaks.
Common questions
How many candidate names should I generate before deciding?
Aim for 30–50 rough candidates before applying filters, then narrow to 10–15 that pass a basic availability and pronounceability check. From there, score your top five against trademark databases and social handle availability. Generating fewer than 20 candidates before deciding often means you chose the best of a weak pool rather than the best option overall.
Is borrowing from another language a trademark risk?
Yes, particularly for common words in European languages that may already be trademarked in international classes. A word that is unused in English may have dozens of trademark registrations in its language of origin. Run any foreign-language root through USPTO TESS and EUIPO's eSearch before committing to it as a brand name.
How do I score candidate names objectively?
Use a simple grid: pronounceability (1–5), spellability after hearing it aloud (1–5), trademark risk (invert: clear = 5, conflict = 0), .com availability (5 = available, 0 = premium-priced aftermarket), and social handle availability (5 = all major platforms clear). The numeric total does not make the decision but forces explicit thinking about each axis rather than letting visual appeal dominate.
What if I love a name but the .com is taken?
Check who owns it and what they are doing with it. A parked domain with no site is negotiable — offer $1,000–$5,000 depending on how generic the name is. An active business using it means the .com is gone, and you should choose a different name or a credible alternative TLD like .io or .co for your category. Do not anchor to a name so hard that you accept a weak TLD for a consumer product.