Digital Minimalism

How many online accounts do you have? We counted 750 million emails to find out (2026)

Published estimates run from five accounts to 168, and most trace back to surveys or password counts. We measured inbox traces instead, and found the answer depends on how deep you scan.

September 9, 2026

15 min read

Copy Link

Quick answer. Yorba members hold an average of 246 online accounts, the figure published in Yorba's 2024 State of Clutter report, drawn from anonymized inbox scans across a member base that has since passed 25,000 people and 750 million analyzed inbox messages. That is higher than every figure anyone has actually measured, and the published record it beats runs from about five accounts to 168. Those other numbers come from asking people to remember their accounts, or from counting saved passwords. We counted differently. We counted accounts that left a trace in an inbox, whether or not the person remembered them. And the number is not fixed, because it rises with how far back you scan.

Last updated September 2026.

Why every published estimate disagrees with every other one
Which widely quoted figures do not survive a fact-check, and what they were originally
The two ways this has always been measured, and why both undercount by design
What an inbox trace counts that a survey and a password vault cannot
Why your account count is a function of scan depth rather than a fixed fact about you
Where the accounts actually live, which is not where you think

Nobody agrees on this number, and the disagreement is the story

Ask the internet how many online accounts you have. You will get a different answer every time.

Not off by a little. Off by an order of magnitude. One page says 26. Another says 300. A third adds up category estimates until it reaches a total no human has ever verified.

We went back to the original source behind every figure we could find. Most of them do not hold.

Source

Figure

Year

What it actually measured

Bitdefender

35.1% say 3 to 5 accounts, 30.9% say 6 to 9

2024

Recall survey, 7,335 consumers across seven countries

Experian

2017

Censuswide survey, 2,001 UK adults

Intel Security

27 logins

c. 2017

Poll of 2,000+ adults. Original release no longer published

Dashlane

150, projected to 300 by 2022

2017

Method not disclosed

NordPass

168 personal passwords

2024

Survey of 1,509 NordPass users

NordPass

120 personal passwords

2026

Survey of 1,509 NordPass users

Yorba

246 accounts

2024

Anonymized inbox scans of member mailboxes

Now the ones that did not make the table, because they should never have been in circulation.

The Statista 90. There is no Statista figure of 90 online accounts, in 2020 or in any year. It appears to be a midpoint of NordPass's 80 and 100 from 2020, washed through a stats roundup until it acquired a source it never had.

The "Google says 27." Both Google and Harris Poll security surveys are real and neither one contains an account count. The 27 belongs to Intel Security, from roughly 2017, and Intel Security's original release is no longer online.

The CNN 100. CNN reported in February 2024 that "each person has an average of 100 online accounts," attributed to NordPass. NordPass's 100 was a count of passwords from October 2020, and by the time CNN published it NordPass had already replaced it with 168. A three-year-old password count became a current account count in one sentence.

That is the state of the record. A survey of British consumers from 2017 sits next to a password company's marketing projection from the same year. A number with no source at all gets syndicated as fact. Figures that counted passwords get quoted as if they counted accounts.

These numbers do not disagree because the world changed between them. They disagree because they were never measuring the same thing.

The two ways everyone has measured this, and why both undercount

There are only two methods behind every figure above. Both fail in the same direction.

A survey measures memory, not accounts

Surveys ask you to count your accounts. That sounds reasonable until you notice the problem. The thing being measured is the thing you cannot remember.

Forgotten accounts are the whole point. The retail signup from 2016. The trial you abandoned. The forum you posted on twice. Ask someone to count those and you are asking them to recall the exact things designed to be forgettable.

Look at what happens when you do. Bitdefender's survey of 7,335 people found the largest group, 35.1%, reporting 3 to 5 online accounts, and only 4.8% reporting ten or more. Almost nobody believes they have more than ten.

Every survey number is a memory ceiling wearing the costume of a total.

A password manager measures diligence, not accounts

Vault counts are better. They are observed rather than self-reported. But they measure a deliberate act, which is saving a login on purpose.

The account you never saved does not appear. The signup you typed once and abandoned leaves no vault entry. So a password manager count tells you how organized someone is.

It also tells you who they are. NordPass's 168 came from a survey of 1,509 NordPass users, which is to say people who had already gone out and bought a password manager. That is not a slice of the population. That is the most organized end of it.

And the number moved. NordPass now publishes 120, down from 168, which it attributes to single sign-on and passkeys. Read that carefully. The number of passwords a diligent person keeps went down. Nothing about that suggests the number of accounts went down with it. Signing in with Google does not mean the account stopped existing. It means it stopped needing a password, and therefore stopped being counted.

The third method: count the traces

There is another way, and nobody markets it because it does not sell a vault.

Accounts leave traces. A registered relationship with a company produces mail. A confirmation. A receipt. A password reset you did not ask for. A policy update. Marketing that arrives for years afterward. Scan an authorized inbox for those traces and you count accounts that exist rather than accounts that are remembered or filed.

An account that emailed you exists whether or not you saved the password, and whether or not you remember signing up. The trace does not care about your memory.

Now the honest part, because research that hides its limits is not research.

This method misses things. It misses accounts tied to email addresses a member never connected. It misses accounts that never sent a single email. It misses anything offline. And Yorba members are self-selected, which is the limit that matters most. They are people who already suspected they had a clutter problem, so they are not a random slice of the population any more than NordPass's users are.

Say that plainly and the number becomes usable instead of inflated. This is the average for people who think something is wrong. It is not a projection onto everyone.

The number

Yorba members hold an average of 246 online accounts.

That figure comes from the 2024 State of Clutter report, aggregated from anonymized inbox scans. It is the highest measured figure published by anyone, and it sits below only Dashlane's 2022 projection, which was a forecast rather than a measurement and which Dashlane no longer cites.

Two things about it are worth more than the number itself.

The first is the gap. People self-report 3 to 5. Password managers count 120 to 168 among their own customers. Inbox traces find 246. Those are not three estimates of one quantity. They are measurements of three different things: what you remember, what you filed, and what exists.

The second is that 246 is a floor, and the reason is the most interesting finding in this entire category.

Your account count is a function of how deep anyone looked

Here is where the question stops having a single answer.

Basic Yorba members authorize two years of inbox history. Premium members authorize up to ten. The 2025 report states the consequence directly: "Majority of Yorba members in 2025 were on a free basic plan, which limits inbox scan depth to the past two years."

Read that against the 246 and something uncomfortable falls out. The highest measured account figure anyone has published was produced mostly by looking at two years of mail.

Two years. Not a digital lifetime. Not the signup from 2013 that still holds your home address. Two years, and it already beats every survey and every vault count on record.

Which reorganizes the whole question. Every published figure is really a statement about how far back the measurer could see. The survey number is what you can see from memory, and memory reaches back about five accounts. The password manager number is what you can see from a vault, and a vault only holds what you deliberately put in it. Ours is what you can see from two years of mail.

So "how many accounts does the average person have" is not answerable as asked. The honest answer is a function rather than a figure, and the variable is scan depth.

Your account count is not a fact about you. It is a fact about how deep anyone bothered to look.

Where the accounts actually come from

An abstract total does not move anyone. So picture where the accounts live.

You would guess the household names. Google, Amazon, the platforms everyone has. They are in there. They are not the story.

In Yorba's 2025 data, the single most common platform appears for only 12% of members. The next appears for 10%, then 9%, and the tenth most common appears for 5%.

Sit with that first number. The most common account any of us hold shows up for barely one member in eight.

So the bulk of your account load is not the familiar logos at all. It is the tail. Hundreds of companies that no two people share. Regional retailers, single-purchase stores, the service you used once in an airport, the forum, the loyalty scheme you joined for a discount code that expired in 2019.

Everyone's tail is different. Everyone's tail is enormous. And a tail made of companies you cannot name is a tail you cannot audit from memory, which is exactly why the survey numbers come back at five.

What the number is doing to you

A large account count is not a trivia stat. It is exposure, spread across companies you cannot name.

In 2025, Yorba members unsubscribed from 52,361 mailing lists, which the report estimates "will prevent over 75,000 future marketing emails annually." That part works. Leaving a list is mostly a solved problem.

In the same year, those members deleted 2,214 accounts.

Look at those two numbers together. Tens of thousands of lists left. A couple of thousand accounts closed.

The mechanism explains the gap. Unsubscribe requests succeeded 84% of the time, because one-click unsubscribe is a standardized mechanism that companies are required to honor. Account deletion succeeded 48% of the time, because deletion is governed by no such standard. One is a button the law forces companies to provide. The other is a request you file and often watch fail quietly. Yorba's deletion team logged 9,996 hours of work in 2025 and lost 1,226 of them to friction alone, which is the time cost of companies making the exit hard on purpose.

So the count keeps climbing while your ability to shrink it lags far behind. Unsubscribing quiets the symptom. The accounts stay open, still holding your name, your address, and whatever else you handed over at signup.

Methodology

This section is not a footnote. It is the reason to publish at all.

Where the 246 comes from. Yorba's 2024 State of Clutter report, which states that "Yorba members have, on average, 246 accounts." The report notes that "all metrics are aggregated from anonymized, system-generated identifiers." It does not publish a sample size or a per-account definition, which is a limit of that edition rather than a defense of it.

The 2025 figures. Everything in this piece labeled 2025 comes from the 2025 State of Clutter report, covering 1 January to 31 December 2025, across 6,979 active members in 132 countries, of whom 4,725 completed a first inbox scan. The 2025 edition does not publish an accounts-per-member figure, which is why the 246 above is dated 2024 rather than presented as current.

How an account is identified. Yorba scans authorized inbox metadata for signup confirmations and automated sends that indicate a registered relationship with a company. The scan reads header data rather than message bodies, runs an initial deep scan of the past two years, then repeats weekly.

Scan depth by plan. Basic members authorize two years of history. Premium members authorize up to ten. Because most members are on the basic tier, these figures reflect recent activity rather than a full digital lifetime, and the account counts should be read as a floor.

What is excluded. Accounts on email addresses a member did not connect. Accounts that never sent email. Anything offline. Every one of those exclusions pushes the true number up rather than down.

Self-selection. Yorba members already suspected they had a digital clutter problem. These figures describe that population, not the general public. The same caveat applies to the password manager figures quoted above, for the same reason.

Comparability. Because methodologies differ across editions, these results should be read as baseline measurements rather than as a year-over-year trend.

Frequently asked questions

How many online accounts does the average person have? Published estimates range from about five self-reported accounts up to 168 counted passwords, and they disagree because they measure different things. Measuring inbox traces instead, Yorba members averaged 246 accounts in 2024. That figure was produced mostly from two years of mail history, so it is a floor rather than a ceiling.

Why do the estimates vary so much? Because surveys count what people remember, password managers count what people saved, and inbox scans count what left a trace. Only the last one catches accounts you have forgotten, and forgotten accounts are most of the total.

Is 246 the same as 246 passwords? No, and the difference matters. A password count only includes logins someone deliberately saved. An account count includes the ones with no saved password, the ones you sign into with Google, and the ones you have not opened since you made them. That is why NordPass's password count fell from 168 to 120 while nobody's account count fell at all.

How can I find out how many accounts I have? Search your inbox for "welcome," "verify your account," and "your receipt," then repeat it on the older address you used to sign up for things a decade ago. It is slow and it misses plenty. Yorba's inbox scan does the same job across your authorized mailbox and puts every account that left a trace in one place.

Do old accounts matter if I never use them? Yes. An account you abandoned still holds your data. If that company is breached, your information goes with it, whether or not you ever logged in again. Unused does not mean harmless. It means forgotten and still exposed. Breach Beacon tells you whether your address already appears in a known breach, in one click and with no signup.

Is a higher number worse? Usually, because every account is a copy of some of your personal data sitting on a server you do not control. More accounts means more places for it to leak and more companies you would have to contact to close the door.

Why can Yorba measure this when other companies cannot? Because the measurement is a byproduct of the product rather than a marketing survey. Yorba scans authorized inboxes to find accounts, so the count comes from the same operation the member asked for, across two plan tiers with different history windows. That structure is what makes it possible to say the number rises with scan depth rather than just asserting a single figure.

Every published number for this question has been a guess dressed as a fact. Built on memory or on vaults, undercounting by design, and in several cases traceable to no original source at all.

The real answer is not a single figure. It is a function of how deep you look, and even the shallowest honest measurement already exceeds every estimate on record.

The practical version is simpler. You have more accounts than you think, most of them belong to companies you cannot name, and unsubscribing does not close a single one of them. If you want to see your own number instead of an average, connect your inbox and let Yorba count the traces. Then start with the accounts that matter most, which are the ones holding your card, your address, and everything else you handed over the day you signed up and forgot.

Less sprawl. More quiet.

Meta data

Meta title How many online accounts do you have? (2026 data)

Meta description Published estimates run from five accounts to 168, and most trace back to surveys or password counts. We measured inbox traces instead, and found the answer depends on how deep you scan.