Typosquatting and domain typos
Two addresses can look alike, one letter apart, without being the same problem. A typo is a typing mistake on a domain that does not exist: it gets corrected. A typosquatted domain exists, belongs to a third party and receives the mail: it gets refused.
Two cases that look alike and have nothing in common
The two cases look alike and have nothing in common. A typo is a mistake by the person, on a domain that does not exist: it gets corrected. A typosquatted domain exists, belongs to a third party and receives the mail: it gets refused.
| Typo | Typosquatted domain | |
|---|---|---|
| Origin | A typing mistake by the person writing their address | Deliberate registration by a third party |
| The domain exists | Usually not | Yes, and it receives mail |
| Server answer | None, there is no server | Favourable, often to every address |
| Course of action | Offer the correction to the person | Discard the address, or decide according to the business policy |
The impersonating domain
A typosquatted domain is a domain registered to look like a well known domain, one letter apart, in order to capture the mail that was meant for it. It belongs to a third party, it really receives mail, and it often answers favourably to every address.
Detection looks up a database of impersonating domains already identified. Each entry names the legitimate domain it imitates, which lets us tell the customer not only that the address is risky, but which domain it is trying to pass itself off as.
An address on an impersonating domain is returned as risky or invalid depending on the customer policy, never as simply valid. The domain receives mail and answers favourably: that is why YesWeCheck spots it through its database of impersonating domains, beyond the dialogue with the server alone.
This is where YesWeCheck goes further than a plain SMTP dialogue. The domain is registered, it declares a server, it accepts mail: every technical measurement is green. What gives it away is its intention, and YesWeCheck reads it in its database of impersonating domains already identified.
Typo correction
A domain typo is a typing mistake by the person writing their address, not a hostile domain. The domain typed usually does not exist, and the intention is easy to read: comparing it with the most widespread mail domains makes the correction obvious.
The correction relies on an edit distance, which counts the letters to add, remove, replace or swap to go from the domain typed to a known domain. The tolerance grows with the length of the domain, and a tie between two candidates produces no suggestion.
| Rule | Behaviour |
|---|---|
| Confidence threshold | No suggestion below 75 out of a hundred. |
| Tolerance | It grows with the length of the domain: a short domain leaves less room for error than a long one. |
| Tie between candidates | No suggestion. Two equally likely corrections are better than the wrong one of the two. |
| Output | The whole corrected address, not just the domain, ready to be offered in a form. |
When confidence passes the threshold, the API returns the whole corrected address, ready to be offered in a form. Below it, the API returns nothing: a wrong suggestion costs more than no suggestion at all.
What to do with it in a form
The correction is offered, never imposed: the person is the one who confirms. A question asked at the right moment, "did you mean…?", recovers the address in one click, without ever changing it behind its author's back.
The impersonating domain, for its part, is not corrected: there is nothing to offer instead, since the person may have written exactly what they meant to write. The course of action is to refuse or to flag, depending on what the sign-up commits to.
Frequently asked questions
- What is email typosquatting?
- A typosquatted domain is a domain registered to look like a well known domain, one letter apart, in order to capture the mail that was meant for it. It belongs to a third party, it really receives mail, and it often answers favourably to every address.
- How does it differ from a domain typo?
- The two cases look alike and have nothing in common. A typo is a mistake by the person, on a domain that does not exist: it gets corrected. A typosquatted domain exists, belongs to a third party and receives the mail: it gets refused.
- How is a typosquatted domain spotted despite a favourable SMTP answer?
- An address on an impersonating domain is returned as risky or invalid depending on the customer policy, never as simply valid. The domain receives mail and answers favourably: that is why YesWeCheck spots it through its database of impersonating domains, beyond the dialogue with the server alone.
- When is a domain correction suggested?
- When confidence passes the threshold, the API returns the whole corrected address, ready to be offered in a form. Below it, the API returns nothing: a wrong suggestion costs more than no suggestion at all.
Go further
Both checks are free and run without SMTP verification. On a sign-up form, they catch two very different causes of loss: a mistyped address that you recover by correcting it, and a hostile address that you discard before it enters your database.
Content updated on 7 September 2026.