Most people pick one username and carry it from site to site. Depending on which platforms you compare, published studies find that between about a quarter and six in ten people keep the exact same username on a second platform, and nearly nine in ten say they have one "main" username.

We read 15 years of academic papers, industry surveys and platform announcements about usernames and put the numbers in one place. This page covers how often usernames are reused, how unique they are, how often they change, and how much a username gives away about the person behind it.

59%of sampled users had the same username across online communities Zafarani & Liu, 2009
89%of people with several usernames say they have one main one Liu et al., 2013
56%of identical-username pairs really were the same person Liu et al., 2013
10%of 8.7 million Twitter users changed username within two months Jain & Kumaraguru, 2014

Key username statistics

  • 59% of users in an early cross-community sample identified themselves with the same username on different sites.[2]
  • 89% of surveyed people who use more than one username prefer to use one main username across communities.[3]
  • People listed on a profile aggregator used an average of 2.44 usernames across 3.92 communities.[3]
  • 26% of Pinterest users who connected a Twitter account had the identical username on both. For Facebook and Pinterest the figure was only 5%.[4]
  • Two accounts with an identical username belonged to the same person in just 56% of checked pairs on travel forums.[3]
  • A study of 10 million usernames found a typical username carries more than 35 bits of identifying information, more than the roughly 33 bits needed to single out one person on Earth.[1]
  • 59% of Twitter handles contain at least part of the person's name.[7]
  • About 10% of 8.7 million monitored Twitter users changed their username in under two months.[5]
  • 1% of Twitter accounts with over a million followers had an old username picked up by someone else within a year.[6]
  • Rare inactive handles on X are priced from $2,500 to more than $1 million.[15]

How many people reuse the same username?

There is no single answer, because the rate depends heavily on which two platforms are compared and how the sample was collected. The studies below all measured the same thing: out of people known to own accounts on two services, how many used the identical username on both.

About 6 in 10 people kept the same username across communities

59% in the 2009 Arizona State University sample

Same usernameDifferent username
Source: Zafarani & Liu 2009 [2], as reported in [3]. 59% rounded to 6 of 10 figures.
Share of people with the identical username on a second platformPublished studies, 2009 and 2013. Each bar is a different sample and platform pair.
Across online communities (2009)59%
Twitter and Pinterest (2013)26%
Facebook and Pinterest (2013)5%

Sources: Zafarani & Liu 2009 [2]; Mittal et al. 2013 [4]. Samples are people who publicly linked their accounts, so they are not representative of all internet users.

StudySampleWhat was comparedIdentical username
Zafarani & Liu, Arizona State University, 2009 [2]38,093 username pairs from BlogCatalog and 36 other communitiesUsername across communities59%
Mittal et al., 2013 [4]328,570 Pinterest users with a connected Twitter accountTwitter and Pinterest26.4%
Mittal et al., 2013 [4]Same usersMore than 90% similar, not identical+10.2%
Mittal et al., 2013 [4]107,910 Pinterest users with a Facebook usernameFacebook and Pinterest5.0%
Mittal et al., 2013 [4]Users with all three accounts connectedIdentical on Twitter, Facebook and Pinterest0.22%

Two patterns stand out. First, reuse is far higher between platforms where the username is the public identity (Twitter, Pinterest) than on Facebook, where most people are known by their real name and rarely set a custom username. Second, exact matches undercount real reuse: on Twitter and Pinterest, another 10% of people used a name that was more than 90% similar, such as the same word with an extra digit.

Reading the range. Taken together, a fair summary is that roughly one quarter to six in ten people keep the exact same username on a second username-based platform, and more do so with a small variation.

Why people reuse usernames

The only published survey we found that asked people directly about their username habits was run by Microsoft Research in 2013 with 153 volunteers.[3]

How people manage usernamesSurvey of 153 people, Liu et al., 2013
Prefer one main username (of those with more than one)89%
Use more than one username in total87%
Use only one to four usernames for everything82%
Keep a main username so there is less to remember80%
Use a single username everywhere13%

Source: Liu, Wang, Zhang, Song, Lin, "What's in a Name? An Unsupervised Approach to Link Users across Communities", WSDM 2013 [3]. Small volunteer sample.

The main reason is memory: about 80% said a main username means fewer names to remember. Respondents also said a consistent name helps them build a reputation that follows them between communities.[3]

The same paper measured behavior as well as opinion. Among 75,472 people who listed their accounts on About.me, the average person appeared on 3.92 communities with 2.44 different usernames.[3] In other words, people needed one new username for roughly every 1.6 accounts (our calculation from those two figures).

The number of accounts per person has grown since then. GWI research published by DataReportal in 2026 finds online adults use an average of 6.75 social platforms each month.[12] More accounts with the same small set of names means more places where one username leads to the same person.

How unique is a username?

The best-known study on this question came from the French research institute INRIA in 2011. The team analyzed about 10 million usernames, 6.5 million from eBay and 3.5 million from public Google profiles, and measured how much identifying information each one carries.[1]

FindingFigure
Usernames analyzed (eBay and Google profiles)about 10 million
Identifying information in a typical usernamemore than 35 bits
Information needed to single out one person in a billionabout 30 bits
Share of usernames below that 30-bit line, across all datasets5% to 25%
Average edit distance between two usernames of the same person4.2 characters
Average edit distance between two usernames of random people12 characters
Matching pairs the method could link while staying the most precise optionup to 71%

A typical username holds more information than it takes to single out one person on Earth

Identifying information, in bits. Each extra bit doubles the number of people a name can tell apart.

Sources: Perito et al. 2011 [1] for the 30-bit and 35-bit figures; 33 bits is our calculation (log₂ of 8 billion). Scale runs from 20 to 45 bits.

People change their username very little between sites

Average number of character edits needed to turn one username into the other

Same person, two sites (example pair from the study)
daniele.perito
d.perito
4.2 edits
average for 10,000 linked pairs
Two random people
mountain_hiker77
bluejay.codes
12 edits
average for 10,000 random pairs
Source: Perito et al. 2011 [1]. The first pair is an example used by the authors; the second pair is invented for illustration.

To put the bit figures in plain terms: picking one person out of 8 billion takes about 33 bits of information (our calculation: log₂ of 8 billion). A typical username in the study carried more than that. Canada's privacy regulator summarized the result as a chance of roughly one in a billion that two people pick the same username by accident.[16]

The study also showed that when people do change their name between sites, they barely change it. Two usernames belonging to the same person differed by about 4 characters on average, compared with 12 for two random people.[1]

Does the same username mean the same person?

Not always, and this is the most misunderstood point about usernames. Rare names almost always belong to one person. Common names do not.

Two accounts, one identical username: same person?

101 hand-checked pairs from travel forums

Source: Liu et al. 2013 [3]. 57 of 101 pairs were the same person.
  • On 53 travel forums with 1.94 million users, researchers hand-checked pairs of accounts that shared an identical username. Only 56% were the same person (57 of 101 pairs they could decide); the other 44% were different people.[3]
  • Those forums contained 325,139 pairs of accounts with a shared username, built from 127,088 distinct usernames.[3]
  • In a Yahoo! Answers sample, 299,716 usernames were shared by 673,037 users, about 2.2 people per shared name (our calculation).[3]
  • After adding a measure of how rare each username is, the same researchers reached 92% accuracy in deciding whether two matching usernames are one person.[3]
  • The most common usernames are simply first names. A 2020 NordPass ranking of the 200 most popular usernames was led by names such as David, Alex, Maria and Anna, with the top entry counted more than 875,000 times.[11]

What this means in practice. A match on "alex" or "maria92" proves very little. A match on an unusual name is strong evidence, but it should still be confirmed with a second detail such as a photo, a location or a linked account before drawing a conclusion about anyone.

What a username reveals about a person

A username is often the only thing visible about an account, and it tends to say more than its owner intended.

What researchers could read from a username aloneThree studies, 2015 to 2025. Different methods and samples.
AI model with web browsing: correct age (2025)90%
AI model with web browsing: correct gender (2025)79%
Text classifier: correct gender from the name string only (2015)74%
Twitter handles that hint at gender (2024)69%
Twitter handles containing part of the person's name (2024)59%

Sources: Alizadeh et al. 2025, preprint, GPT-o3 given only the X handle of 1,384 survey participants [9]; Jaech & Ostendorf 2015, 44,000 dating-site usernames [8]; Islam & Finlayson, ICWSM 2024 [7].

Anatomy of a username

Share of handle parts by type, from 500 Twitter handles split into 1,213 parts. The handle shown is invented.

sara
37.7%first or last name
k
4.5%initial
1994
8.3%number: year, age or date
nyc
1.9%location
Source: Islam & Finlayson, ICWSM 2024 [7], Table 1. Other parts were sentiment words (3.4%), acronyms (1.8%) and parts with no clear meaning (42.5%).
  • Researchers at Florida International University built a parser for Twitter handles and reported that 59% contain at least part of a person's name and over 69% indicate gender in some way.[7]
  • When they broke 500 handles into parts, 37.7% of the parts were first or last names, 8.3% were numbers such as a year, age or date, 4.5% were initials and 1.9% were locations.[7]
  • A University of Washington study predicted gender from the username string alone with about 74% accuracy on 44,000 dating-site usernames (a 25.8% error rate).[8]
  • In a 2025 preprint, an AI model with web browsing was given nothing but the X handles of 1,384 survey participants. It matched their self-reported age with 90% accuracy and gender with 79% accuracy.[9] The authors recommend that AI providers restrict this ability in public products.

Username changes and recycling

Usernames feel permanent, but on several platforms they are not. When a name changes, old links and mentions can end up pointing at a different person.

How often people change username

1 in 10 Twitter users changed username in two months

853,827 of 8.7 million monitored accounts, October to November 2013

Changed usernameKept the same username
Source: Jain & Kumaraguru 2014 [5].
  • A study that monitored 8.7 million Twitter accounts for two months in 2013 found 853,827 of them, about 10%, changed username in that time.[5]
  • In a closer look at 2,698 users who changed their name, 45% changed it more than once, and one account changed 96 times in 117 days.[5]
  • About 61% of changes happened within 24 hours of the previous change.[5]
  • 36% of repeat changers went back to a username they had dropped earlier.[5]
  • Around 75% of changes produced a name that looked quite different from the old one; only about 10% were small edits.[5]

What happens to the old username

  • In a sample of 70.8 million Twitter accounts, researchers found 106,935 usernames that had been used by more than one account, involving 196,200 accounts (0.27% of those observed).[6]
  • Among 4,263 accounts with over a million followers, 42 (about 1%) had a former username taken by another account within a year. Twenty of those 42 takeovers broke Twitter's rules and were suspended.[6]
  • 11.3% of the reused usernames still had links pointing to them from other websites.[6]
  • At the time of that 2017 study, Twitter, Pinterest and Tumblr released old usernames for anyone to claim, while Facebook and LinkedIn allowed changes but did not hand the old name to someone else.[6] Platform rules change, so check current help pages before relying on this.

When a platform changes everyone's username

In 2023 Discord dropped its four-digit tags and moved every account to a unique username. The company said more than 40% of users did not remember their tag or know what it was.[13] Discord reported 150 million monthly active users at the time.[14] For anyone searching by an old Discord name, that means a large share of pre-2023 names no longer match the account they once did.

Usernames as login names

Reuse matters for security as well as privacy, because on many services the public username is also half of the login.

  • A 2010 Trusteer study of more than 4 million browser users found that when a bank lets customers choose their own user ID, 65% use that same ID on non-financial websites. Even when the bank assigns the ID, 42% reuse it elsewhere.[10]
  • In the same study, 47% reused both their banking username and password on other sites.[10]
  • An IBM survey of 22,000 people in 22 markets found the average person created 15 new online accounts during the pandemic, and 82% reused credentials at least some of the time.[17]

The market for usernames

Short and common usernames ran out years ago, which has turned them into something people pay for.

  • In October 2025 X announced a Handle Marketplace for inactive usernames. Rare handles, with X giving @Pizza, @Tom and @One as examples, are priced from $2,500 to more than seven figures.[15]
  • Longer inactive handles such as full names are offered to top-tier subscribers at no extra charge, but return to the pool if the subscription ends.[15]
  • When Discord planned its 2023 switch, it gave the first pick of new unique usernames to the oldest accounts, because not everyone could keep the name they wanted.[13]

Timeline: 15 years of username research and platform changes

  1. 2009First cross-community study finds 59% of users keep the same username [2]
  2. 2011INRIA shows usernames alone can link profiles, using 10 million names [1]
  3. 2013First survey of username habits: 89% have a main username [3]
  4. 201410% of 8.7 million Twitter users found to change username in two months [5]
  5. 2017Over 100,000 recycled Twitter usernames documented [6]
  6. 2023Discord moves all accounts to unique usernames [13]
  7. 202459% of Twitter handles found to contain part of a real name [7]
  8. 2025X starts selling inactive handles for $2,500 to over $1 million [15]; AI models shown to profile users from a handle [9]

What the numbers mean for you

  • If you want your accounts kept apart, use a different username for each one that matters, and avoid small variations. The research shows names that differ by a few characters are still easy to connect.
  • Leave your real name, birth year and city out of any username you want to stay private. Those are the parts researchers and software read first.
  • If you change a username, remember that on some platforms the old one can be claimed by someone else, along with every old link and mention.
  • If you are checking on someone else, treat a matching username as a lead, not proof. For common names it is wrong almost half the time.

Method and limits

This page is a review of published work. We searched academic archives and publisher sites for studies that report a measured number about usernames, then read each paper and took the figures from the paper itself where it was available. Industry surveys are included only where the company named its sample size.

  • Many of the core studies are old. The largest reuse studies date from 2009 to 2017. Few researchers have repeated them since platforms restricted data access, which is itself a finding: there is no recent large-scale measurement of username reuse.
  • The samples are not representative. Reuse studies rely on people who publicly linked their accounts, who are likely more consistent than average. The 2013 survey had 153 volunteers.
  • Studies measure different things. The rates in the first chart come from different platforms and years, so they show a range, not a trend.
  • One source is a preprint. The 2025 AI study [9] had not been peer reviewed when we checked.
  • Our own calculations are labeled in the text. They are simple arithmetic on the published figures.
Cite this page

You are welcome to quote these statistics. Please credit the original study for each number and link here as the collection.

Reverse Username Search. "Username Reuse Statistics 2026." Published November 14, 2026. https://reverseusernamesearch.com/username-reuse-statistics/

Frequently asked questions

What percentage of people use the same username on different platforms?

Published studies put it between about 26% and 59%, depending on the platforms compared. A 2009 study found 59% across online communities, and a 2013 study found 26% between Twitter and Pinterest. In a 2013 survey, 89% of people with more than one username said they have a main one they prefer.

Does the same username always mean the same person?

No. On travel forums, two accounts with an identical username were the same person in only 56% of the checked pairs. Rare usernames usually point to one person; common ones such as first names are shared by many.

How often do people change their username?

In a study of 8.7 million Twitter accounts, about 10% changed username within two months, and 45% of those who changed did so more than once.

What can a username reveal about a person?

Often a name and a likely gender. A 2024 study found 59% of Twitter handles contain part of the person's name and over 69% hint at gender.

Can someone else take my old username?

On some platforms, yes. A 2017 study found over 100,000 Twitter usernames that had passed from one account to another, including former names of accounts with more than a million followers.

Sources

  1. Perito, Castelluccia, Kaafar, Manils. "How Unique and Traceable are Usernames?" Privacy Enhancing Technologies Symposium, 2011. arxiv.org/abs/1101.5578
  2. Zafarani, Liu. "Connecting Corresponding Identities across Communities." ICWSM, 2009. ojs.aaai.org. The 59% figure is as reported in [3].
  3. Liu, Wang, Zhang, Song, Lin. "What's in a Name? An Unsupervised Approach to Link Users across Communities." WSDM, 2013. microsoft.com/research
  4. Mittal, Gupta, Dewan, Kumaraguru. "The Pin-Bang Theory: Discovering the Pinterest World." 2013. arxiv.org/abs/1307.4952
  5. Jain, Kumaraguru. "@I to @Me: An Anatomy of Username Changing Behavior on Twitter." 2014. arxiv.org/abs/1405.6539
  6. Mariconti, Onaolapo, Ahmad, Nikiforou, Egele, Nikiforakis, Stringhini. "What's in a Name? Understanding Profile Name Reuse on Twitter." WWW, 2017. arxiv.org/abs/1702.04256
  7. Islam, Finlayson. "A Semantic Interpreter for Social Media Handles." ICWSM, 2024. cs.fiu.edu
  8. Jaech, Ostendorf. "What Your Username Says About You." 2015. arxiv.org/abs/1507.02045
  9. Alizadeh, Gilardi, Samei, Mosleh. "Web-Browsing LLMs Can Access Social Media Profiles and Infer User Demographics." Preprint, 2025. arxiv.org/abs/2507.12372
  10. Trusteer, "Reused Login Credentials," 2010, as recorded by PasswordResearch.com. passwordresearch.com
  11. NordPass, most popular usernames, 2020, as reported by SecurityBrief. securitybrief.co.uk
  12. DataReportal, "Digital 2026: 2 in 3 people on Earth now use social media." datareportal.com
  13. Discord, "Evolving Usernames on Discord," May 2023. discord.com/blog/usernames. The 40% figure is as reported by the Associated Press [14].
  14. Associated Press via Fortune, "Discord forces members to change usernames," May 9, 2023. fortune.com
  15. TechCrunch, "X to launch marketplace for buying inactive handles," October 20, 2025. techcrunch.com
  16. Office of the Privacy Commissioner of Canada, blog post on username linkability, February 16, 2011. priv.gc.ca
  17. IBM Security and Morning Consult consumer survey, 2021, as reported by Help Net Security. helpnetsecurity.com