Discuss: Sign Up Forms Must Die

by Luke Wroblewski

51 Nice Article but misses some issues....

It is indeed a nice article and clearly demonstrates the bugging issue of Signup forms. But it misses some issues –

1.)A better approach to kill signup forms is OPENID. This article has not mentioned it at all.

2.)Second thing is, when I tried to comment on this article, I was served a “Create an account “ form first and then comment later. I think Alistapart.com should itself try to implement
what has been said in this article.

posted at 04:45 pm on March 28, 2008 by Nilesh Walmik

52 Re: nice article but misses some issues...

Nilesh, I disagree with the practice of using OpenID for anything other than a blog – so I don’t think it is really relevant in this article. Centralizing our information and consolidating it may make things seem simple or easy, but really what is your identity on the internet? If we hold all this info in one place, it could be compromised more easily and used against use in a more horrific way. This scares me because it is like giving out an ID card – something a person shouldn’t need on the internet but I fear it will happen soon anyways.

I agree with the idea of this article though. It is annoying to have to give up so much of your info just to see how something works. I would like to apply this in some of the projects I have in my future.

posted at 04:53 pm on March 28, 2008 by John Josef

53 What about COPPA?

This seems like a great idea, but collecting personal data without age verification (13+) is illegal in the US. Also, having played with geni a little bit, quite a few people have felt like they’ve been spammed, since there’s little indication that they’ve actually signed up for anything.

posted at 03:30 am on March 29, 2008 by Adam Freetly

54 Short, Single Sign On, & Record sets

As folks continue to bring up OpenID, I’ll re-iterate: single sign on solutions also have the potential to remove the need for sign-up forms. However, they have yet to become widely adopted by the majority of Web users (hopefully this changes over time), and they only resolve the issue of user identification. Single Sign-on solutions don’t let people experience applications and services before requiring them to authenticate.

The point here is that the initial experience with a Web-based product can be used to engage new customers and illustrate what a service offers and how it works. User authentication doesn;t meet that goal.

Similarly, keeping sign-up forms short (while a very worthwhile pursuit) doesn’t address this point either. The fundamental problem is we have a tendency to treat people as record in a database. That’s what the sign-up form is: an output of the fields in a recordset that uniquely identify a user. This is not how people think of themselves and their relationship to product or service. So while we can make sign-up forms shorter and we can try to spread single-on sign solutions, the bottom line is we are concerned first with the recordset, and second with the person coming to our site or service.

I genuinely feel there is an opportunity to do better.

posted at 05:08 am on March 29, 2008 by Luke Wroblewski

55 OpenID != Single Sign On

@John Josef and others:

With respect, I think you need to understand more about OpenID before you accuse it of “centralising” information. OpenID is a decentralised identity system, which is why I and many others use it. I did not, and never would, use a single sign on system for anything other than a closed network for precisely the reasons you state.

posted at 07:17 pm on March 30, 2008 by Jonathan Baker-Bates

56 Another way to look at it?

I don’t think there’s a one size fits all about when to present the user with registration. The comments to the article suggest this – some prefer to sign up straight away, some would rather do it last.

Luke is spot on when he keeps bringing us back to understanding the user’s goal. I’m sure he would agree that the best way is to base your design on a distillation of both the user’s needs AND those of the business. Balancing those needs should yield a form which: – has not too many fields for the user and not too few for the business, and – doesn’t ask for registration too early for the user but doesn’t miss out on valuable information about the user.
In my model of the 4 Cs of good form design" – which maybe of interest to some readers – it is essentially a problem of context" and conciseness.

Also, as noted by a number of commenters, the problem is not necessarily that sign up is required but rather that the site doesn’t communicate well the with user. Expectations should be managed through clear information about what the application can and can’t do, what is and isn’t available with registration and what registration information is used for.

The fourth “C”, cleverness , rears its head as well. If a site remembers the things you’ve done while testing it, then registration can be relatively seamless.

So the solution is maybe not to kill off ALL web forms but rather to use a model like the 4 Cs to ensure they are well designed, just like any other form!

posted at 06:58 am on March 31, 2008 by Jessica Enders

57 I agree

that signing up for things stinks in the long run. Just now I signed up to post on this site, and I had to fill out another form. Some of the information you give is for demographics, and others are just plan stupid. I was going to sign up for an AOL account and they asked me to supply my phone # and address. I stopped right there because I know that if I finished who knows what kind of stuff I would get in the mail from AOL.

posted at 10:48 pm on March 31, 2008 by Sean Goff

58 Irony Alert.

I wanted to comment this, and after attempting I actually had to sign up. Hahahaha. I was going to say, I’ve seen sites lately where I don’t need to sign up, but it asks for just my name and a link. I would prefer all sites like that.

posted at 03:13 am on April 1, 2008 by Addison Kowalski

59 Does Registration Elevate the Discourse?

Great excerpt, great discussion – I look forward to reading Luke’s book. My comment is, as Jeffrey Zeldman pointed out earlier, slightly off-topic as it relates to commenting/posting rather than testing out web services, but since the registration issue keeps coming up, I thought I’d toss this in to the discussion.

I am currently doing some ad hoc research around the effects of registration on the volume and relative value of user-generated content. I agree with others here that the inclusion of a minor registration barrier should elevate and enrich the discussion – at least that’s always been my hunch – but I was recently referred to the following article that suggests that registration actually dilutes quality discourse:

http://blog.topix.com/archives/000106.html

Has anyone else here witnessed the phenomenon described here, where the absence of a registration barrier actually generates more and higher quality responses from users? If this is true, then the only benefit of the registration barrier benefits the business (potentially less spam to comb through).

The vast majority of premium content sites that allow commenting and voting ask for, at minimum, a registration via email address and a password before users can submit content. But does the fact that everyone’s doing so make it a best practice? We strive in all other ways to reduce the amount of administrative debris we present to users, but in this case we seem to agree that it’s worthwhile to derail them from the completion of their task.

posted at 06:24 am on April 1, 2008 by Matthew Derby

60 RE: RE: The Irony

Registering before you can comment helps encourage a higher level of conversation and discourage (although not eliminate) spam. For these reasons, such registration has become best practice and industry standard on sites that allow readers to post comments.

Although a common, I wouldn’t call it best practice. This doesn’t bring any value to the users who just want to post one comment, it can even discourage from posting. The way this site does it is to make users enter a username and password, but without requiring an email address for confirmation. This doesn’t stop spam, and doesn’t stop multiple registrations, so why not just let users comment directly, and only require to enter a “type this word” function (with an image showing scrambled letters) to prevent bot automatic posting?

posted at 07:17 am on April 1, 2008 by Friend of Users