Company Email Formats: The Patterns That Cover Most Businesses

Most business email addresses follow one of three patterns: first.last@ (jane.smith), flast@ (jsmith) or first@ (jane). One provider's July 2026 analysis of 336,782 work emails found those three in 82.6% of addresses combined. To use a pattern, learn the company's format from one real address, apply it to the name, then confirm with an SMTP check.

This guide covers the patterns, how they change with company size, and a worked example of confirming a guess against a real mail server, including the two ways it goes wrong.

The patterns, with how common each is

first.last@ is the most common format, but the order varies by company size. The best recent public data comes from Sendburg's analysis of 336,782 profiles in its own B2B database, run on 16 July 2026. It is one vendor's view of its own data, not a census, but the method is published: each address was matched against the person's first and last name.

Pattern	Example for Jane Smith	Share of addresses
first.last	jane.smith@	47.71%
flast	jsmith@	26.81%
other / unmatched	—	8.57%
first	jane@	8.13%
firstlast	janesmith@	2.29%
first_last	jane_smith@	2.29%
f.last	j.smith@	2.14%
last	smith@	1.20%
last.first	smith.jane@	0.65%
first.l	jane.s@	0.13%
first-last	jane-smith@	0.10%

The top three cover 82.65% of addresses. The top six named formats cover 89.37%. In practice that means a finder only needs to try a handful of candidates per person, and the work is in confirming which one is real.

Company size changes the odds

Bigger companies are far more likely to use first.last@. From the same dataset:

Company headcount	Share using first.last@
1–10	38.0%
11–50	36.3%
51–200	45.2%
201–500	47.3%
501–1,000	52.8%
1,001–5,000	63.2%
5,001–10,000	68.2%
10,001+	74.2%

The reason is practical. At ten people, jane@ is unique and friendly. At ten thousand, there are several Janes, and first.last with a rule for duplicates is the only scheme that scales.

What small businesses actually publish

On small-business websites, the named addresses we found were mostly first@. On 1 October 2026 we checked 337 roofing, plumbing, HVAC and electrical businesses in 11 US metro areas (sourced from OpenStreetMap; method in our local business list guide). 111 published an address on their own domain. Most of those were role addresses like info@ and office@. Only 19 published any named person's address, and those 25 addresses broke down as:

Format	Addresses	Businesses
first@ (pete@)	21	15
flast@ (jsmith@)	4	4
first.last@	0	0

It is a small sample and only covers addresses people chose to publish, so read it as direction, not proportion. The direction matches the size table: at owner-run firms, try first@ before first.last@.

How to find a company's format

One real address from the company tells you its pattern. Sources, in rough order of usefulness:

1. The company website: team, about, contact, careers pages and the footer.
2. A search for the domain: "@company.com" in quotes finds addresses in press releases, conference pages, published papers and directories.
3. Email signatures in any thread you have with anyone there.
4. Press releases and investor relations pages, which usually name a media contact.

Two checks before you use what you find:

• Is the email domain the same as the website domain? Not always. In a separate check of 60 large US companies, four had no mail server on their main website domain at all because they receive mail on another one (for example jpmchase.com rather than jpmorganchase.com). Use the domain from a real address.
• Is the address current? A published address can outlive the person. Our worked example below includes one that no longer exists.

Confirming a guess: a worked example

A verifier asks the company's mail server whether it would accept mail for an address, then hangs up before sending anything. The answer to the RCPT TO command is the evidence. We ran these checks on 2 October 2026 (UTC) against two real small businesses. Domains and names are shortened here; the full transcript is in our research file.

Case 1: a roofing company on Google Workspace

The site publishes pete@. We asked about that, the formal version of the name, and a random string:

RCPT TO:<pete@[roofing-co].com>
250 2.1.5 OK

RCPT TO:<peter@[roofing-co].com>
550-5.1.1 The email account that you tried to reach does not exist.

RCPT TO:<zzq7x2kd9w0p1v3c@[roofing-co].com>
550-5.1.1 The email account that you tried to reach does not exist.

This is the case where verification works. The server rejects a random string, so its 250 for pete@ means something. And it shows why nicknames matter: peter@ is not an alias. A guess based on the formal name from a licence record or LinkedIn profile would have bounced.

Case 2: an HVAC company on its own mail server

The site publishes mmorgan@, which suggests the flast pattern. We checked it and four variations:

RCPT TO:<mmorgan@[hvac-co].com>     550 No Such User Here
RCPT TO:<m.morgan@[hvac-co].com>    550 No Such User Here
RCPT TO:<morgan@[hvac-co].com>      550 No Such User Here
RCPT TO:<morganm@[hvac-co].com>     550 No Such User Here
RCPT TO:<morgan.m@[hvac-co].com>    550 No Such User Here
RCPT TO:<zzq7x2kd9w0p1v3c@[hvac-co].com>  550 No Such User Here

The published address does not exist. The server is clearly answering honestly, since it rejects the random string too, and it rejects the address on the company's own website. The likeliest explanation is that the person left and the website was never updated. Without the check, that address would have gone straight into a sequence and bounced.

Case 3: the one you cannot confirm

If a server accepts a random string, it will accept every guess, and none of them can be confirmed. That is a catch-all domain, and it is not rare. In the same 337-business sample, of the 145 mail servers that gave a definite answer, 93 accepted a made-up address. On those domains, pete@, peter@ and asdfgh@ all look identical from outside.

An honest finder or verifier returns those as risky, not valid. Send to them only with other evidence, such as the address being published, and in a small separate batch.

Edge cases that break simple patterns

Real names do not always map cleanly onto a pattern. The usual traps:

• Nicknames: Pete or Peter, Bill or William, Kate or Katherine. Try the name people actually use, which is often the one on the website or in LinkedIn's display name.
• Duplicates: the second John Smith may be john.smith2@, jsmith2@ or john.a.smith@. Large companies have their own rule.
• Hyphens and particles: Mary Smith-Jones, Ana de la Cruz. Companies variously keep the hyphen, drop it, or use only one surname.
• Accents and apostrophes: José or Jose, O'Brien or obrien. Most companies strip them.
• Case: RFC 5321 technically allows mailbox names to be case-sensitive, but says that "exploiting the case sensitivity of mailbox local-parts impedes interoperability and is discouraged". In practice, write addresses in lower case.

The process, in order

1. Find one real address at the company and note its domain and format.
2. Write the candidate for your person using that format and the name they actually use.
3. Verify it. Valid: send. Invalid: try the next most likely pattern for that company size. Risky: decide on other evidence.
4. Never mail every permutation to see what sticks. It damages your sender reputation and, in the US, addresses generated by permuting names are treated as a harvesting problem under CAN-SPAM.

You can run single checks with our free email verifier, or let an email finder try the patterns and verify them for you.

The takeaway

Three patterns cover most of business email, and company size tells you which to try first: first.last@ for large companies, first@ for owner-run ones. But the pattern is only half the job. Confirm every guess against the mail server, expect some published addresses to be stale, and treat catch-all domains as unconfirmable rather than found.