People who want to trick your customers or staff rarely use your real domain. They register one that looks close enough at a glance: a missing letter, two letters swapped, a look-alike character from another alphabet, or your name under a different ending. This tool lists those lookalikes for free and, when you ask, checks which ones someone has actually registered.
Two steps: list, then check
Show variants builds the list on our servers. Nothing is looked up, so it is free and instant. You see how many lookalikes your name produces, how many of each kind, and the most similar ones first, with the characters that differ from yours highlighted.
Check which are registered turns the list into a job. For each of the most similar variants (200 by default, up to 500) we ask the ending's own nameservers whether the name is delegated. Registered ones are then looked at more closely: their mail servers, their web addresses, the registration date from the registry, and the title of their homepage. This step needs a free account and costs 1 credit per 10 variants checked.
Kinds of lookalike
- Typos: a missing, doubled, swapped or extra letter, a slip to the next key on the keyboard, a different vowel, an added hyphen, or a dot that turns part of your name into a subdomain of another.
- Look-alike characters: rn for m, vv for w, 0 for o, and letters from accented Latin, Cyrillic or Greek that are drawn almost like yours. Such names are registered as punycode, the xn-- form you see under each one. When every letter of your name has a twin in one alphabet, the whole name can be rewritten in it, which is the hardest kind to spot.
- Other endings: your name under .net, .co, .shop and two dozen other common endings.
- Bit flips: a letter changed by a single-bit memory error in someone's device. Rare per visitor, but a popular domain gets enough traffic that some of it lands on these names.
Every name is checked against the rules for valid internationalized names, and mixed-alphabet labels are left out, because browsers show those as punycode anyway.
How the risk rating works
Each registered lookalike gets a score from 0 to 100, from four signals:
- Registered recently: within 30 days counts most, then 90 days, then a year. A lookalike bought last week is more likely to be part of something happening now.
- Can receive email: it has mail servers that accept mail (a null MX doesn't count). That is what phishing and invoice fraud need.
- Very similar: look-alike characters score highest, then swapped and missing letters.
- Serves a website: its homepage answers. We fetch only the homepage, show its title as plain text, and never follow its links.
A score of 60 or more is high risk, 35 to 59 medium. Registration dates are looked up for the riskiest 50, from each registry's RDAP or WHOIS service, within that registry's rate limits; for some endings the registry publishes no date at all.
Reading the result
The list is sorted by risk, so the ones to deal with first are at the top. Filter it to the names that can receive email or were registered recently. Each row shows the kind of lookalike, the registrar, the mail servers and what the homepage answered. Names that didn't answer clearly are listed apart; running the check again later usually settles them.
A partial result means the check ran out of time before reaching every variant. You pay only for the variants that were checked.
What to do next
Block high-risk lookalikes in your mail gateway, make sure your own domain enforces DMARC so nobody can send as you, and report abuse to the registrar shown in the row. The linked guide walks through each step, including the dispute procedures open to trademark owners. To see who holds a name, use the WHOIS lookup; to see how a look-alike character is encoded, use the punycode converter.
Privacy
Your usage history keeps the domain you checked and nothing about the variants. Lookalike results stay in your account: they can't be published as shared reports.