You sat down on a Saturday with a clear vision and one perfect company name. Then you typed it into a registrar and watched the registrar suggest "yourname.dance". The .com is parked at $14,000. The .io is taken too. The .ai wants $9,000.
Welcome to founding a company in 2026, where the seven-letter dictionary .com namespace was exhausted around the time of the dot-com bubble and the two-word combinations went years ago. Here is the founder playbook — seven moves that actually work, with the cost, perception, and SEO trade-offs that nobody tells you up front.
Before you reach for an alternative, find out which kind of "taken" you're dealing with. There are four flavors and the move is different for each:
| Status | What it means | Move |
|---|---|---|
| Active business | Real site, products, traffic | Pick a different name. Trademark risk if you push. |
| Parked / for sale | Auction listing or "this domain is for sale" page | Negotiate (see #6) or skip. |
| Privately held, dormant | Resolves to nothing or a placeholder, no listing | WHOIS lookup + cold email. Lower offers work. |
| Pending drop | Registered but expiring within 30 days | Backorder via DropCatch / SnapNames / NameJet. Often $69 if no one else bids. |
Confusing the four costs founders thousands. A "premium $4,800" domain marked for sale is often a dormant registration the owner never touched — they grabbed 600 of them in 2014 and pay $13/year on autorenew. A direct cold email beats their listing 30% of the time.
The single most reliable move in 2026 is the prefix or suffix word. get, use, try, hey, join, app, hq, and your category itself (labs, cloud, ai) all open up large pools of available .coms.
Worked examples that scaled past $100M in revenue with this exact pattern: Stripe → never needed it, lucky. Notion → grabbed a six-letter dictionary word, also lucky. But: GetMagic, UseGather, TryRamp, JoinHandshake, AppLovin. The "verb + brand" pattern is so normalized in 2026 that buyers don't even register it as a workaround anymore.
Cost: Standard .com registration, ~$10–13/year.
SEO: No penalty. The brand becomes the search query within months.
Risk: Pick a verb that doesn't conflict with the natural sentence ("get" + verbs feels weird; "get" + nouns is great). And do the trademark check on the full string, not just the brand part.
If you can stomach a 30-second "wait, how do you spell that" cost in the first year, coining a new word is the most defensible long-term play. Notion, Tumblr, Plaid, Brex, Vercel, Linear — all coined or cropped real words to claim a clean .com namespace nobody else can ever take.
The 2026 brandable name framework that works:
The fastest way to find candidates is to generate hundreds and filter aggressively. The TinyTools generator produces 100 brandable candidates in 8 seconds, with availability checks built in, so you can throw out 95 and study the remaining five.
This used to be controversial. In 2026 it isn't, for specific TLDs in specific categories:
.ai for AI/ML companies — fully normalized, often preferred over .com by AI-native founders..io for developer tools — still standard, no perception cost in dev audiences..app for mobile-first products — Google-owned, requires HTTPS, signals "this is software.".dev for developer infrastructure — Google-owned, HTTPS-required, signals technical product.What to avoid in 2026: .xyz, .co, .tech, .cloud, .online, .shop. Consumer-facing audiences still pattern-match these as "couldn't get the real one." If you're selling to non-technical buyers, the perception cost is real and measurable. We've seen A/B tests on landing pages with .com vs. .xyz versions of the same brand show 8–15% conversion drops on the alt-TLD variant.
Cost: .ai ~$70–90/year, .io ~$45/year, .app/.dev ~$15/year.
SEO: Search engines treat all TLDs equally for ranking. Branded queries are unaffected. Direct-type traffic is the only risk — buyers who try .com first land on a competitor or a parked page.
Mitigation: Buy the .com if you can ($1k–$10k for a 6-letter dictionary word) and 301-redirect it to your .ai. Best of both.
The classic ccTLD hacks: bit.ly, last.fm, about.me, del.icio.us. They worked because the brand spells out cleanly and the TLD is short.
2026 candidates that are still open:
.so (Somalia) — works for verbs and adjectives. ~$30/year..gg (Guernsey) — gaming-coded. ~$70/year..fyi, .cc, .me, .tv — viable for the right brand..is (Iceland) — common verb hack ("get.is", "what.is"). ~$50/year.The catch: ccTLDs can be revoked at the country's discretion. .io is technically the British Indian Ocean Territory, and ICANN's 2024 review of decolonization-related TLDs created real uncertainty about its long-term future. Use a ccTLD as your vanity URL or product subdomain, not your primary corporate domain — and always own the .com equivalent if you can, even if it's just for redirect.
The 2010s startup move: Tumblr, Flickr, Lyft, Grindr, Toggl, Scribd, Wufoo. Drop a vowel from a real word and you have a clean .com nobody else can take.
This still works in 2026 if you avoid the lazy patterns. What's burned out: dropping the final "e" or "er" from English verbs (overdone). What still works: compressing two short words into one (Mailchimp → six syllables down to two), cropping a Latin or Greek root (Vercel from "verstehen" / "vert"), or adding an unexpected letter (Pinterest from "pin" + "interest").
Risks: SEO direct-type traffic loss is real ("I'll just type it...wait, was it Tumblr or Tumbler?"), and the founder-and-team explanation cost is permanent. Mitigate with a clean wordmark logo and the redirect from the misspelled .com if it's available cheap.
If the .com is parked at a number, you have three doors:
We wrote a longer field guide on the negotiation game: How to Negotiate Buying a Premium Domain Name in 2026. Short version: the more emotional you are about the name, the worse the deal you'll get.
The move founders resist hardest is the one that almost always pays off. Spending three weeks negotiating with a domain squatter, or three months explaining a vowel-dropped misspelling to every customer, or three years redirecting traffic from the .com a competitor owns — none of that is worth it when the actual answer is "the name wasn't that special, find a better one."
"We had a name we loved, the .com was $40k. We tortured ourselves for two months. Eventually picked a coined name we love MORE that cost $12. The original sounds like a stretch in retrospect." — A founder, on a thread we won't link to
Names you fall in love with on day one aren't predictive. Most founders rank their final name higher than their original favorite within 90 days of launch. Generate 200 alternatives, score them blind on a Friday, and you'll find one cleaner than what you started with.
If a real business owns it: skip. If it's parked at $50k+ and you don't have funding: skip. If it's a category-perfect .com at $1–5k and you have product-market fit signals: buy. If you're pre-seed or pre-product: pick a brandable, generate 200 candidates, ship the cheapest one that passes the dictation test. The brand is built by the product, not the domain.
.com ("get-yourname.com"). High direct-type loss to the unhyphenated competitor. Almost never worth it.| Extension | Year-1 cost | Renewal | Notes |
|---|---|---|---|
.com | $10–13 | $10–15 | Default. Negotiable on premiums. |
.ai | $70–90 | $70–90 | Two-year minimum at most registrars. |
.io | $35–50 | $35–50 | Status uncertainty, but still common. |
.app / .dev | $13–18 | $13–18 | HTTPS required. Google-operated. |
.so | $25–40 | $25–40 | Verb hacks, vanity URLs. |
.gg | $60–80 | $60–80 | Gaming/community brands. |
For the registrar itself, we use Namecheap (they handle all of the above and don't upsell privacy as a paid extra — it's free with every domain). They're an affiliate partner of TinyTools, so the link in our generator results gives you the same price they list, with a tiny kickback to keep the tools free.