5 ms·
These are all really just UX-focussed _improvements_ to make it easier for a user to fill out forms. From a _users_ perspective - well, at least my personal per
by ghoul2 15y ago
These are all really just UX-focussed _improvements_ to make it easier for a user to fill out forms. From a _users_ perspective - well, at least my personal perspective - the reasons are a lot different:
1. I am not convinced you need the information you are asking me to fill out - especially if I am just trying to figure out what your service does. Generate "guest" credentials for me if you want me to try it out. I am really not going to bother registering at the dozens of new services I try out (or attempt to try out) every week. Not worth it. Once I have taken it for a test drive and am convinced the service if of value to me, I will gladly fill out the information needed, but yet again, will refuse to provide even a single bit more than I consider _strictly_ necessary to provide the service. For example, If you are providing say, email alerts using text messages, I do not see why you need to have my A/S/L.
2. I do not believe it when a new service "promises" to keep my information private and secure, or to never spam me. This promise is rarely kept - maybe it was meant when it was made, but soon the service gets new investors, maybe a new marketing head who believes _her_ new email newsletter is NOT spam, etc etc. You know the best way to have that promise kept? To not share the info in the first place!
3. For the same reason, I will not give you access to my FB/Twitter credentials by using their SSO solution to sign-in to your service. That's Too Much Information. SSO - BrowserID or whatever, is not a solution in this context. Its merely a way to reduce the impact of a problem which shouldn't exist in the first place!
4. Each form is preventing me from doing what I wanted to do - try out your offering. Decide what is more important to you - that I try out your product, or that you get another row filled out in your marketing database?
5. Sure, a well designed form is better than a badly designed form. Whats even better? No form at all. Imagine a brick and mortar retail location wishing to interview me regarding my personal info, buying habits etc before they allow me to enter and browse. Never happens. Instead they try to entice you later, once you have actually decided to purchase something, to signup for their loyalty card to gain that additional info.
- moe 15y ago[...] Generate "guest" credentials for me if you want me to try it out. This can not be stressed enough. Have a working live-demo, accessible with a single-click. Not screenshots. Not screencast. And for god's sake not "Schedule a Webinar". Live demo, period. I can't count the number of times I've been mildly interested in a service but couldn't be bothered to fill out anything just to see how it really feels. Stripe.com has it right. They do offer registration but you're free to "Skip this step" and with that one click you're dropped right into the live-dashboard to play around and explore.
- minhajuddin 15y agoI built a small blogging engine http://substancehq.com http://substancehq.com, and have had more than 30 throwaway blogs created by "generated guest credentials" on the first day. This is a great thing for a new guy like me. However, most users don't follow through and finish the sign up, which tells me that the app itself is not appealing to the users. This approach is a lot better than having bounces off your signup form. I wonder why many people don't do this.
- nchlswu 15y ago> Sure, a well designed form is better than a badly designed form. Whats even better? No form at all When it comes to /true/ UX, I think this is pretty much it. Forms aren't necessary in a hell of a lot of times and often forms are reduced to reduce friction (with the goal seemingly to increase conversions) then ask for all true information later; I don't see the point. As for the article, as you mentioned these are UX focused improvements. Just because forms have been designed with these improvements in mind, it doesn't mean these are actual factors the user is considering. I'm not saying this isn't true, but articles like these bother me a bit because they inferences and not fact as implied.
- StavrosK 15y agoWhy won't you sign in with Twitter? The best scenario for me is if they only require read-only access to your timeline. They don't get your email (I think), and they can't do anything to your account. It's pretty much the least amount of information you can give, since all the Tweets are public anyway.
- bartonfink 15y agoPossible reason - lack of a Twitter account. I don't have one and have little interest in ever obtaining one as Twitter doesn't solve any problem that I have. I could create a Twitter account and only use it as proxy credentials, but that seems like needless complexity.
- Domenic_S 15y ago> Twitter doesn't solve any problem that I have. Sounds like you contradicted yourself...
- r00fus 15y agoHow? I'll just create a password manager entry for the site, just like any other site. If I'm asked random silly required questions, I'll lie if I don't think it benefits me in some way or is truly required. Most users at this point (not desiring external auth) would just give up unless they saw a huge need for the service.
- bartonfink 15y agoIf Twitter is marketing itself as an authorization service, then I am contradicting myself. It doesn't do that, however, and using it as a proxy for that is an unnecessary hoop for me to jump through.
- ghoul2 15y ago1. Its not simply about they not being able to "do anything" to my account. Its the linkage. Maybe I use my real name on twitter, and I do not wish to use it with the service I am trying out (or the otherway round). The linkage extends even further. If I signin to both, service A and service B with the same twitter login, now these two can link to each other, without any assistance. 2. It takes mental effort to keep track of all the permission I may have granted to various services from my various identity providers. If I want to disable access, I need to make extra effort. If there is no linkage, I merely forget it about it and it goes away. Even you say "They don't get your email (I think)" - that "I think" right there is the problem. I would have to keep (and maintain) a mental checklist of everything I am providing access to when I use a syndicated login service. 3. Permissions evolve. Twitter may change its access policies, a new hack might allow privilege escalation. Each time this happens, I would need to review what apps have which permissions. For services I actively use, I am willing to put in that effort If absolutely needed, but for something I tried out as an evaluation, its not worth the hassle. As I said, until I am convinced I am going to be using the service, I do not wish it to stick around in my "apps with permissions" list, or anything else for that matter. If I try it out, and decide it aint for me, I just want to navigate away and never think about it again.
- cletus 15y ago> I am not convinced you need the information you are asking me to fill out This. I hate overly intrusive sites. If I'm buying something from you, you need to know who I am, fine. If you won't let me download the binary distribution of a GPLed software package without entering my name, email address and date of birth, then you get fake information. I'll go to the point of using a throwaway email address if I have to. Back 10+ years ago Amazon was one of the first sites that allowed you to shop and do everything without even registering right up until the point where you placed the order and made payment. At the time, most sites wouldn't even give you a working shopping cart without registering. Amazon figured this out in the 90s. I don't know why sites persist in increasing the friction of attracting new users. I can only imagine it comes down to: 1. Being easier, from a development point of view; and 2. Misguided marketing wonks. Either way, cut that shit out. Seriously. And the same goes for only supporting Facebook/Twitter login.
- toyg 15y agoYou forget: 3. Monetizing through shady data-reselling channels. Who cares if your data is not 100% correct because the occasional smart nerd will use a throwaway email? Somebody somewhere will pay something for the database and empirically prune it later on their own. Especially for "free" services without a clear monetization path, that must be a strong temptation.