Trust and personal information security are very important issues these days when it comes to online accounts, so it is not unusual for people not wanting to create an account (i.e. provide personal information) when making purchases or requesting information.

So it would seem like most cases when users are forced to do so it is due to a business or technical rather than user requirement.

So the question is: what are the design patterns or techniques that creates a smoother user experience when forcing users to sign up or create an account? Bonus points if these techniques are actually ethical design practices as well!

  • 67
    Who's to say we need to require people to create an account when making purchases? Commented Oct 14, 2018 at 23:27
  • 2
    You can use something like Stripe and even manage them as customers through the api. Commented Oct 15, 2018 at 1:26
  • 33
    Unless customers are buying content say like Netflix , I see no need to ever require creating an account. If I ever come across a site that requires that I sign in, I skip it and go somewhere else.
    – Michael J.
    Commented Oct 15, 2018 at 15:06
  • 8
    @ESR: But that's assuming that you are going to be making repeat purchases, and frequently enough so that info like credit card numbers/expirations (or even your address) are going to stay current between them. This isn't always the case.
    – jamesqf
    Commented Oct 15, 2018 at 18:13
  • 18
    forcing users to make an account is missing from the title, the obvious answer to which is: don't.
    – Mazura
    Commented Oct 16, 2018 at 18:18

12 Answers 12


Update: i have extended the answer and added some examples, i might further enhance the quality of the answer if i get feedback

Thought you'd never ask, here are some things to look at:

1. Does the user even need to create an account? Why?

Many product managers/designers fail to ask themselves this important question, it has become one of the largest trends that every digital product out there needs to identify its users, while it's true we are much more able to achieve things when users create an account, it has to be a conscious decision rather than the intuitive ofcourse. By forcing everyone to create an account you are making an authority bet, which in many cases you lose.

Login Walls

This is a very good article on nngroup's website called Login Walls Stop Users in Their Tracks it explains a lot about the impact of such walls.

2. Ask only for the relevant information for the current task

When the user is creating an account, the user has something in mind, a need to use your product/a task to be achieved, this requirement is merely an interruption to what the user is seeking, it's crucial that we keep the amount of user input as minimal as possible.

LinkedIn Sign Up

No intention to promote linkedin, but consider their case as most of us know the platform. Eventually we - the users - enter so much information for our profile, can you imagine if linkedin asked for all of that at the registration phase?

And yes believe it, some do ask a lot! The largest platform i know of that made a record of such mistake at its release was Google+ - RIP, we had to go through so many forms some to reflect our social aspect, some our professional, some interests etc.. It took me personally around 20 minutes to get past the initiation process.

3. Ask for the right information at the right time

Imagine having to do your lifetime laundry at once, not so fun instead we prefer to do them on a weekly basis, user input is a similar case on a much smaller scale - considering the exaggeration it's more like microscopic scale. but! it gives the same effect.

Rather than asking for the address at the registration before even trying your product ask it on the checkout, or the more important one should the user register before using your product or at the very end of it.

Create Account After Checkout

PS: This is mistake is common at the hands of fullstack developers since if the mindset used for designing the database is the same as designing the forms, lord's mercy.. the poor user..

4. Simplify the User Input

Go as creative as it gets with this one because there are so many ways to approach it a couple of examples:

4.1. Single Sign On, Social Logins and the one click registration Social Login

4.2. Fill the form on behalf of the user and let the user correct it when possible

Ever wondered about Facebook's default birthdate values for signing up? Facebook Birthdate

Day and Month clearly represent today, but why 1993? maybe as of this year the most common sign ups are users born in 1993?

I'm sure many of us here know we can guess the country from the IP address, what else can we guess is [email protected] implying First Name Michael & Last Name Lai?


This is an example of how tolerance could result with a slightly enhanced experience, this case does not apply 100% of the time, the question is if we are able to research patterns used in emails and found that x% of the users use [email protected] then we could try to use that information in auto filling the form even if it doesn't work all the time.

In this example i have intended to break the rule to prove that it still can workout somehow, on a mobile device the user would only have to click the button in the middle to swap the values between first name and last name, also the user would have to press backspace 4 times eventually requiring the user of 5 taps of barely thinking over having to type in manually everything.

Use this for mobile apps, web forms already provide autofill for users that might become useless if you fill it.

Disclaimer: this example was made within minutes it is not verified please do not take it as is to your product without testing it, it's merely about the approach and how to think about such problems.

4.3. Question the fields you're asking for, it reflects on the interaction pattern, do i need to ask for the full birthday? or am i only concerned in knowing the user is over 13 years old? the difference is Date Field vs Checkbox i guess you know how much that is in user time/frustration.

4.4. Field grouping and sequence, reflect the user mental model.

4.5 Don't exaggerate password complexity (unless you have a good reason)

4.6 by @vlaz: do i need to ask for the full birthday?

"perhaps it's some sort of statistical anomaly but a lot of services that ask for that information attract disproportionate amount of people born on the 1st of January."

thank you @vlaz

and the list goes on..

  • 34
    do i need to ask for the full birthday? perhaps it's some sort of statistical anomaly but a lot of services that ask for that information attract disproportionate amount of people born on the 1st of January.
    – VLAZ
    Commented Oct 15, 2018 at 6:21
  • 4
    +1 plus tell the user the advantages to them of having an account with you. If you can't come up with any advantages, don't demand an account.
    – Paul Smith
    Commented Oct 16, 2018 at 10:32
  • 3
    guess the country from the IP address NO NO NO Use the HTTP headers for that, because they show the user desired language and not an increasingly inaccurate location. May I also remind you that different countries can have different languages ?
    – Antzi
    Commented Oct 19, 2018 at 3:25
  • 2
    @Antzi Amen to that! Assuming that one country maps to one language is stupid - I speak English and live in a French-speaking region, but because the country happens to be Switzerland, PayPal sends everything to me in German... Commented Oct 22, 2018 at 6:58
  • 5
    I can only imagine the face of the filling-form guy who has indeed been born on Oct 18 1993.
    – Wtower
    Commented Oct 22, 2018 at 12:46

Interesting question. To answer it I think we have to understand what's frustrating to users and then provide some alternate ways of doing things.

Also remember, if you don't need the user to create an account then don't make them!

Lengthy Forms

Users may find it frustrating to fill out a bunch of information in many different input fields. Especially when most of the information isn't going to be used. Try to only collect information you absolutely need and try to collect it in the most efficient way possible. Allow the user to take many shortcuts. While it may take more time and be quite annoying to include these shortcuts, it will make UE much better. For example, if an address field needs to be filled out, instead of making the user type their full address, city, state, country, zip code, etc. have the form automatically search as the user types in the address and fill in the information automatically. Also, have common domains autofill for emails (@gmail.com, @hotmail.com, 65+? @aol.com).

Passwords, passwords, passwords

This is a big one. Nobody likes passwords, nobody can remember passwords, and everyone hates them. Their approval rating is literally 0%. But, unfortunately, we need some way to authenticate a user. So, simplify this as best you can. Don't put extreme requirements (7.5 characters (Yeah you gotta figure out how to type 0.5 of a character), Uppercase, lowercase, numbers, can't have consecutive numbers, must include a Chinese character, etc.) these are extremely annoying to users and many will screw it up a bunch of times before getting it correct. Also, that isn't the best security practice either.

Allowing the user the option to login with another service (Google, Facebook, etc.) is a good option.

Additionally, when you sign up, you should automatically be signed in. Countless websites do this stupid thing where you signup and then are forced to retype your email, username, password just to sign back in.

Lastly, if the user does screw something up, don't clear everything out making the user retype all their information (ehem client side validation).


Many websites will require verification in some form of a text or an email. Many times emails will be delayed making the user wait extended periods of time. Allow the user to continue with what they were doing BEFORE verification. Make the last step confirming the email/text, giving time for the user to receive it.

Spamming my email

Simple: Don't do it. If the user wants to buy your product they will go to your site and purchase it, you don't need to send them constant reminders "[RANDOM ITEM THE USER DOESN'T WANT] 50% OFF LIMITED TIME". If you want to send the user emails about discounts have the user tell you they want that. Either have a subscribe button on item pages or something else.

Don't share their information

Don't be Facebook.

  • 33
    "Countless website do this stupid thing where you signup and then are forced to retype your email, username, password just to sign back in." or they don't redirect you to the page you were previously on before having logged in. This is UX 101.
    – ESR
    Commented Oct 15, 2018 at 4:33
  • 13
    Site-managed autocomplete for anything besides physical-address lookup is likely to be a UX disaster. Properly specifying input type and autocomplete attributes is substantially friendlier. Commented Oct 15, 2018 at 7:59
  • 41
    Add "don't block password managers" to the list
    – Ferrybig
    Commented Oct 15, 2018 at 14:13
  • 9
    Re autofill: No, NO, NO! Unless it's something like automatically looking up the state from the ZIP code, few things are more annoying.
    – jamesqf
    Commented Oct 15, 2018 at 18:15
  • 9
    Ph and make sure your password verification matches the rules shown to the user (and show them!) I had a website complain that my 8 character password wasn't a minimum of 8 characters once. Or just today one said that my "any three types" (upper, lower, number, symbol) password that contained upper, lower, and symbols wasn't valid (because it didn't have a number). Commented Oct 15, 2018 at 21:34

I'm sorry for the presumption since I know nothing about your use case, but I'll be the one to say it: don't force them to make an account. Just don't. Users can understand when making an account is necessary. If they expect to use your site without an account, then either you've confused them about what the site does or it's probably possible.

That strong opinion stated, there are some options that work essentially like an account, but don't feel to the user like they've created an account. After an interaction, you can email them a unique link or code (for example inside a "confirmation" email). They can use that link or code to access relevant information in the future (e.g. shipping information). Another thing you can do, though I haven't seen it in the wild, is offer a "retrieve my info" page where they enter an email address and the link/code gets resent to them. Basically this is like logging in via a password reset email every time they visit the site.

Where is the line between the above tactic and creating an account? I would view this as ethical as long as you do not collect and store information about them or personalize their experience. If a user has not "signed up for an account" then I think they have some expectation of privacy regarding what is being collected and associated with their email address.

  • 1
    1st paragraph, +1. But if you have my email then I'm essentially already signed up. If it's a fake email, what was the point? That most people are idiots and will use their real email? You have my permission to presume all you want because this question belongs on IPS; it's a people problem, not a computer problem, and besides that it has a duplicitous title.
    – Mazura
    Commented Oct 16, 2018 at 23:23
  • 2
    @Mazura, I definitely see where you're coming from. A more polite way of putting it is probably: if a user might rationally prefer to give a fake email, then providing an email and/or making an account should absolutely be optional. In the rest of my post I'm assuming the user has some good/useful reason for giving their email.
    – usul
    Commented Oct 17, 2018 at 1:45
  • +1. I'd also add that it is OK to ask user "Do you want account with that" after he did what he needed to do - like, filled shipping form. After. Optionally. When he already seen that all the data collected was really needed for what he was about to do anyway.
    – Mołot
    Commented Oct 18, 2018 at 14:14

Single Sign-On

Heavily utilize SSO services.

Add as much support for as many platforms as possible. Even though this is time consuming and compatibility issues increase as you increase the number you support, look at it like you are supporting more platforms (Android, iOS, Windows, Linux) because odds are the user will have one type of single sign on account that they regularly use.

Generate Passwords By Default

Another practice I've seen is generating the password for the user while still giving them the option to change the password if wanted. If the user wants to come back, then they'll write it down or change it. Odds are if they don't, it won't even bother them. You have their email, and if they want to get a password reset it would already be as easy as having access to their email/phone in most situations, so there is no actual loss of security as long as the password is generated correctly and sent of HTTPS.

Don't Ask Them for A Username

Lastly, don't make them come up with a username and email, one is enough.

If you follow all of these, you'll just need to make the purchase form ask for their first, last name (etc). This will allow for the creation of an account to happen almost transparently, and the user, beyond accepting the terms and conditions and clicking create, will not be hassled. One could even argue that the account is more securely created when the user doesn't have to create their own password, as they will not be prone to reuse a password from another site.

  • 2
    +1 I guess this works when there is a 'trusted' network of sites that the user is comfortable with exchanging or sharing information on. However, if the user wants to remain anonymous then what would be the best strategy? And will the users be fully aware of the implications of providing such information of SSO?
    – Michael Lai
    Commented Oct 15, 2018 at 4:33
  • 10
    don't make them come up with a username and email, one is enough I'd even say that username isn't always needed. If it's not going to be displayed (e.g., having public posting), then an email is enough. Users usually remember their email, so they don't need to remember another identification. If there is a need for username, by the gods, don't make it set in stone at the time of account creation. There is no real reason to.
    – VLAZ
    Commented Oct 15, 2018 at 6:17
  • 6
    For what it's worth I refuse to use SSOs (e.g., Google, Facebook, etc.) to login or create an account on any website. You'll get my junk email address, nothing more.
    – ohmu
    Commented Oct 15, 2018 at 20:43
  • 3
    @DavidKamer I have yet to find a service compelling enough for me to create a junk SSO account.
    – ohmu
    Commented Oct 15, 2018 at 21:57
  • 3
    @DavidKamer but if you already have a gmail account signed in to google services that you do want, setting up a junk gmail account for SSO is one misclick away from linking your primary (and potentially very personal) online ID. I use google/SSO for a few things that started via Android so it was almost necessary (e.g. Strava) but otherwise an email address I control
    – Chris H
    Commented Oct 16, 2018 at 15:49

As you mention online purchases, the best is to let people checkout without needing to creating an account. They still have to fill out tons of details for their delivery, just ask them at the end if they want to create an account so they can check on their order easier.

Doing this, you ask it at the right time, when there is a benefit for the user.

  • And if they do, auto-populate as much information as you can from the order they just placed (don't ask for their address again). Commented Oct 15, 2018 at 21:45
  • Actually, when you do ask to create an account, there should be no need to ask about anything except law-required permissions and maybe username & password.
    – Mołot
    Commented Oct 18, 2018 at 14:17

Use Social Sign Up

You are right creating a new account requires considerable effort and 86% of surveyed people report being bothered by it. However, it has been found that social sign up increases conversion rates by around 52%.

It is not hard to explain why. Social signup requires only one click and you're in. Compared to the traditional way where you need to decide which email and password to use, recall the last password you used, and deal with password rules which could force you to invent new password. It just requires a lot of cognitive load to recall, store and remember passwords.

A very good example of using social sign up is Pinterest

enter image description here

Here you don't even have to click if you are already logged in to Facebook or Google. The registration is virtually frictionless. As a plus, 67% of your consumers are more likely to return to your website because you don't force them to create username or password. Social sign up also increases user engagement.

enter image description here

The most popular social login options are Facebook and Google. In addition, there is a comprehensive study outlining the most popular social sign up options by business vertical.


Among the other proposed methods for reducing user registration pains we should consider social sign up as it shows promising results. It requires considerably less effort to sign up, and increases the user engagement or your returning users.

  • 15
    Please, don't make these services the only option. I'm a happy email-and-password user myself, and I don't want to connect any other accounts (and in fact, I have no accounts on places supporting this).
    – SilverWolf
    Commented Oct 16, 2018 at 2:45
  • 1
    @SilverWolf I didn't mean social sign up only. Commented Oct 16, 2018 at 6:58
  • 1
    Sorry, wasn't trying to say you did. Just leaving this here in case someone else saw your answer and thought "Gee, that sounds great! Let's remove email and password, because nobody uses it!"
    – SilverWolf
    Commented Oct 16, 2018 at 11:29
  • 3
    Facebook is bad enough, but using it to login everywhere is a security risk that needs to be managed. Commented Oct 18, 2018 at 18:30

There is a whole new approach, based on the Ethereum Blockchain.

It is called "Universal Logins". The base idea is to create a key pair and deploy a contract on the Ethereum Blockchain that accepts messages from that public key.

The goal here is not to have the best onboarding experience on the ethereum ecosystem, but the best login anywhere on the internet. For starters the current running code has these advantadges:

  • No need to type or remember a password anywhere
  • Instant login in multiple devices
  • No need to download or install anything extra
  • No single server with private data that can be attacked or leaked (but beware of the public data you share on the blockchain)
  • The user can take the account they created in one app and use it to login in another app
  • If the app goes offline, the users can still access their data
  • User is in control

Drawbacks are, creating the user itself has a cost, since their account is a contract that needs to be deployed.

More Details here: https://medium.com/@avsa/universal-logins-first-demo-1dc8b17a8de7

  • +1 Thanks for your contribution to UXSE. I have looked into some of the concepts in the blockchain space around personal identity and information security and it is certainly full of innovation!
    – Michael Lai
    Commented Oct 17, 2018 at 3:49
  • 2
    While very technically interesting, I'm not sure this will improve the user-experience for creating a user-account. It has advantages in consistent procedure, but convincing the user to spend the time doing it and to trust the info-sec of the site is the focus of the question as I understand it. Commented Oct 19, 2018 at 13:53

1. Highlight the benefits

“You do not need to create an account to make purchases on our site. Simply click Continue to proceed to checkout. To make your future purchases even faster, you can create an account during checkout.”

Source: The $300 Million Button

2. Create a sandbox

When authentication is necessary (i.e. for reading and writing data etc.) offer the user a temporary guest account to try it out and kick the tires in order to make sure it works for them before creating their own permanent account.

3. Hire a doorman

Imagine a Doorman at a fancy apartment building saying "Welcome home Ms. Kimball" and letting you in. Wouldn't it be great if technology could just recognize you while requiring additional things from strangers who want entry?

Apple's FaceID is an example of "Doorman authentication" which may eventually carry over into authenticating online. This would virtually eliminate the need to create accounts and remember passwords all over the place while introducing some other problems such as one person filling the role of multiple users.

Further explained in the answer below...

4. Trust is hard to gain and easy to lose

Storing personal things about a user without their knowledge will erode trust. Don't store personally identifiable information without first asking permission.


Patterns Without Accounts

It is possible to create a rich user experience without forcing users to create an account. Three prominent examples: PUBG Mobile, Imgur, Microsoft Office.

Guest Accounts: PUBG Mobile offers guest accounts. Users are identified by a globally unique identifier and are able to progress through the full gameplay experience. Accounts are opt-in and provide a better experience by unlocking multi-device syncing and cloud backup. If a user invests enough time in PUBG Mobile to require those features, they’re likely to want an account (vs feeling forced). I don’t know if in-app purchases can be made without an account. Personal anecdote: I tried PUBG Mobile instead of Fortnite (account required) because PUBG made it easy to start playing without an account.

Throwaway Accounts: Imgur allows anonymous uploads. I don’t think they can be edited once the browser session ends. Additional community features open up to people who choose to create accounts (it seems like they may be gradually shifting to requiring accounts as usage grows, but I don’t use it enough to know for sure).

Data Files: Microsoft Office provides tools for interacting with data, but stores the data on the user’s computer. The Microsoft 365 version of Excel adds syncing and other cloud features in exchange for creating an account. A number of applications implement this approach by storing documents in a user’s Dropbox account rather than on their computer.

Matchmaking: I’m unaware of a notable example, but WebRTC chat apps offer a less common example, which is that a user can go to the site, generate a one time token, then provide that token to someone else to initiate a chat. The token may or may not persist beyond a single use.

I hope the above help you to think of some ways you might be able to let people use your app without an account. As these examples show it is possible to defer account creation until users demand an account from you (or even forever).

  • I encountered a craft supply website last year with ephemeral accounts. You could opt to sign in or continue without sign in. Without sign in no user information required to make it to the checkout, then during the checkout process your payment, email, and shipping info was taken but you were never given a username and password to the site. Only the specifically required info was taken and only at the time it was required. I really appreciated how smooth the process was.
    – Myles
    Commented Oct 19, 2018 at 14:47

To the premise: business absolutely wants users to create an account. Eg to hold personal information or some such.

It's how you collect the information that matters.

Passive registration allows you to collect the data as part of a purchase procedure without the user burdened by creating an account. At the end of a purchase the user can be asked if they want an account created based on information they have voluntary supplied.

If user chooses to create account: you have fulfilled the business need without forced registration

if chooses not to create account: you probably break GDPR and the user wouldn't have gone through a forced registration anyway


This is about e-commerce shopping sites.
SaaS products usually require an account before any meaningful work can begin.

Dump the misconception

First, let's get rid of the biggest misconception in this discussion (not just here, but across the e-comm and usability inudstries):

Security is not the issue

Think about it: Whether I create an account or not, I'm going to authorize you to collect payment from me somehow. That's the biggest security risk in the whole transaction and there's no way around it. Paypal and other similar third party handlers have tried to mitigate that risk to a large extent, but I'm still going to have to connect you to that service somehow.

Account saturation is the problem

When a user is confronted with the need to create yet another account, they immediately think about all the other accounts they already have rattling around in their head. Most people don't use password managers, so the thought of adding another account to the pile just to complete a purchase is exhausting, annoying, and sometimes overwhelming.

Make account creation optional

This is the first step toward a healthier check-out funnel. No matter how easy (or secure) you make your account creation, some people just don't want to bother. And you don't need them to. Just let them roll through check-out entering all the necessary information for shipping and billing and don't ask for anything more.

Let the user join at the end of checkout

Put on your reality glasses: The user provided almost everything needed for an account during check-out! Most of the time, the only missing piece is a password (or social account connection).

If you explain the benefit to repeat visitors (easy repeat ordering) and deal shoppers (sale notifications) and the fact that you already have almost everything you need, most customers will be happy to fill that last field.

Put the Create my account button and password field right there at the end of checkout, and you'll be surprised how many people who opted out earlier dive right in.

Provide social sign-on

When the whole O-auth thing hit the scene years ago, few people realized how it would change the business. The implication of security was there: keeping your key account details with an org focused on securing them. And that was primarily the value prop.

But when we started running user research on it over time, the biggest take away was that users just wanted to think about one sign-in everywhere. The truth is, using your Facebook account to sign-in to e-comm sites is, in most cases, less secure than creating a scoped account for a certain site. But users didn't understand or care about that issue — they just wanted less accounts to remember!

Providing social sign-on has become a requirement over the years. Not everyone wants it, but most people do. Add that to your end-of-checkout sign-up process and you'll get a lot of new accounts.


No passwords! At all!

I would like to add to previous, excellent answers that it's possible to create a login-site which neither demands nor store passwords.

Use case

This is what it would look like when the user wants to login on a password-less site...

  1. The user enters his/her email adress.
  2. An email with a login link is sent (much like when one resets a password on a traditional site).
  3. The user follows the link.
  4. The website uses the login token in the link to login the user.
  5. The user is kept logged in for X days. The login link expires after a few minutes.
  6. After X days the user is automatically logged out and need to login again.

Basically a "Reset password" function used as login.


  • No password for the user to remember.
  • No password stored in the server that can be stolen.
  • Can be combined with two-factor-authentication.


  • A new way to login which the user might not be familiar with.
  • E-mails might be delayed making a login taking longer than desired.
  • Not compatible with password managers.

Further reading

  • 3
    This is an awful login experience. There might be delays with message delivery. Message can go to spam. The process pulls user out of context: they were on some web site, they clicked the login link, now they have to go to whatever they use to read their email (it might be just another tab in the browser, but it might be a beast of a program, think Outlook) and click something there.
    – n0rd
    Commented Oct 15, 2018 at 23:10
  • 4
    This is awful for mobile users. I have to open the browser/app, click the login button, switch to the email app and click the link in a way that doesn't make it open inside the email app. It probably is more secure than the email/password combo, but it's a big hassle for the user.
    – user118333
    Commented Oct 16, 2018 at 9:42
  • 3
    @Morfildur Isn't that a little bit like saying two-factor-authentication is awful because you need to switch to the code generator app/look att your text messages on the mobile? I get the impression that you are magnifying problems that doesn't need to be problems. Commented Oct 16, 2018 at 9:56
  • 2
    -1 "No password to be stolen" People couldn't care less if there passwords were stolen, people care because that password protects stuff that they don't want to be accessed. You are still storing information that can be stolen. Commented Oct 17, 2018 at 14:09
  • 4
    @BjörnLarsson The user experience of two-factor authentication is awful. I'll generally only tolerate it when the consequences of having the account in question compromised are more awful.
    – jmbpiano
    Commented Oct 17, 2018 at 15:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.