[Usability] What users really "want" ?
Abhishek Parolkar
Satish N Kota
Hi Abhishek,
Nice to see that we finally also came out with the Usability issues… I agree with you that though this is nothing to do with Rails it is to do on how to build your websites… I also agree rails have an edge over PHP on Usability and AJAXification…
Looking at the A/B test, I see you have had more conversions to do as you had prepopulated more details. However my question what is the minimum number of entities required to signup a person? One/two/three or anynumber. Many times I have wondered what is that is required…I feel the best way to get more conversions is to ask only one thing which is email address…which also acts as your login…Then send an email with an auto generated password… which they can use and authenticate themselves…and in the first sign on ask them to change their password…
I always have understood that emotions act on when people are trying to signup for activities… I have debated quite a lot with my clients on this … but yes…culturally and location wise it changes…I asked an US Client what they suggest, they wanted user to provide first name, last name, email address(two times), password(2 times) and Captcha key… and finally authenticating via a link when they receive an email… Great we go for it…but when I suggested it for a german client he threw it out…as he said… in germany if people are asked so many questions then they will not put in until and unless they have emotionally made the decision… so the basic thing is to identify what actual information is required to sign up a user…
However for any people who are concerned with Usability should sign up with Human Factors International (HFI) news letter… They help you in giving you an understanding of how your usability should be to get your clients converted…I used to follow them some time back…so I am not sure about their link… I think its www.hfi.com / www.humanfactors.com
Thanks and Regards
Satish N Kota
Kumar Gaurav
> Hello,
>
> I am writing my first non-ruby post in the list :-) because I feel this might be helpful for the geeky audience who are in middle of web app development world and some of us do think about "usability" when we code, because Rails developers are considered "Pleasant Web" developers ;-).
>
>
>
Abhishek,
What I am writing here is more of a user feedback on whol.ly sign up
process than anything else. But one can draw some conclusions from it.
So I did try signing up on whol.ly
The first thing I noticed was why do I have to click multiple times to
get to where I live? I for one would definitely prefer a text box
where i can quickly say jayanagar and the map should zoom to jayanagar
and then you can expect me to pin point my exact address. So what I am
saying is, don't give me lot of form fields in the first step. Just
give me one. And let me decide how I want to navigate to my address.
By clicking or by typing.
Here is an idea.
Have the map on the first page like it is. Provide one single form
field on top the map for me to type in my location ( more thought
given on this later ). Also have the address form next to the map
where you have blank space right now. This is the setup.
Now this is what i will really really like: I clicking on the map and
you filling up the address real time. And then I can correct the
address if I want.
Now regarding that form field on top of the map. That may or may not
be redundant in this setup. More thought can be given on this.
Another small point. I prefer the hybrid map rather than the street
map. I think it is more personal. And since you are banking on people
to play around with the map, I think defaulting the map to hybrid may
not be a bad idea.
Again YMMV. And it was good of you to share the results.
cheers
p.s : I did not complete the sign up process. I am a bit finicky about
my House Number being up there on the net. I do understand what you
guys trying to do. I am still wary. But then I am like that all the
time. :)
--
Kumar Gaurav
http://www.kumargaurav.info
Latest Post -- Gitting things right
http://wp.me/pbU2Q-2j
sharath ruby
Great Abhishek
Most of us we simply release the page once our friends & colleagues say it is good. Thank you for pointing more scientific & systematic approach. Thank you.
Sharath
From: bangal...@googlegroups.com [mailto:bangal...@googlegroups.com] On Behalf Of Abhishek Parolkar
Sent: 22 November 2009 16:08
To: bangal...@googlegroups.com
Subject: [ Bangalore RUG]
[Usability] What users really "want" ?
Hello,
Amit Mathur
Great finding and thanks for sharing with us.
I suppose, user registration is the biggest bottleneck applications
face with regard to adoption. For the "site visitor" -> "registration"
-> "paid customer" funnel, it is common to have only 10% success at
each stage. I know that even a very large e-commerce site has less
than 30% success at registration (i.e. > 70% cart abandonment).
One other idea that is gaining ground these days is where you don't
ask users to register, but let them use the site as if they were. Then
allow them to be a registered user once they are convinced of its
value. Of course, its not suitable for all applications but yours may
be suitable. Say, let them choose their location on map but don't ask
for email, name etc upfront. See here for more details:
http://www.alistapart.com/articles/signupforms
Regards,
Amit.
AbhishekParolkar
Your inputs are immensely valuable , Thanks for it.
All,
We are improving the user experience of whol.ly from time to time
and you will see lot more improvements and results of our tests
available openly. It will be injustice to this thread if I further
monopolize this thread by talking only about whol.ly as an
example :-).
In course of my career in conceptualizing , building, promoting &
maintaining various web apps. I have come across few surprises about
user behavior that I would love to share with people. ( may be I can
list them in bullets)
1.) "Users dont know, what they want" : Initial feedback loop in
startups consist of close friends or usability consultants, who bring
in few users to evaluate the system , Everybody comes-in with mindset
to beta test the application. They look at each feature, they look at
each screen and make suggestion about features/look-n-feel/faults.
*surprisingly* nobody asks the "tough" question,"Does the product
makes any sense?" , It takes an army of completely unknown people who
come to website to explore something. Only their behavior reveals the
truth. Big Takeaway: Dont ask users what they require, just ask whats
their "difficulty" , put forward a solution and observe their
acceptance without them knowing about it.
2.) "Eyecandy is essential but surprisingly, not vital": We all, new
generation web 2.0 freak think that a pleasurable interface is all we
require to make users keep coming back. No doubt! , a fancy interface
impresses people, they come and say "Wow , Thats pretty cool". But
when you do the numbers to figureout how many users actually came back
to use the site, truth is revealed. It always take 3 times more than
your own time estimations to create that right look-n-feel. Takeaway:
Dont bother about "how your text field/button looks?" until you have
figured out that people are ready to fill-up a form. (or atleast
create a right balance of eyecandy vs usability/changeability of your
stuff)
3.) "They demand verbosity first & then they want you to shut up":
Demands of a new user turns out to be completely different when same
user becomes "returning user". They first need an interface that
guides them with a conversation, and when they know what they want to
do with the app, they want you to "shut" your instructions *off* and
let them find shortcuts to get things done. Takeaway: Dont build for
returning users unless you have new users, and when you have users
that starts returning back dont bore them with same interface.
These are my observations (not the rules), and I am still
experimenting and validating other facts. Let me know if you have any
other observations. Surprisingly, I admit, My own views in the past
were very different from these observations.
- Abhishek Parolkar
- http://github.com/parolkar
- http://twitter.com/parolkar
sharath ruby
Also from users to know what is not required helps at times.
Regards
Sharath
-----Original Message-----
From: bangal...@googlegroups.com [mailto:bangal...@googlegroups.com]
On Behalf Of AbhishekParolkar
Saurabh Bhatia
A friend of mine and me had a similar discussion on captcha and how to eradicate a step like that. Less Reverse Captacha is a step in that direction, http://github.com/stevenbristol/less-reverse-captcha/ , it is running in production for us, and till now it has been successful in keeping the spam away.
The Usability of the app depends on a host of factors :
1. Who the users are ? - General Users Do not remember many usernames, I prefer a solution like https://rpxnow.com/ which makes it dead easy to login using any id that user has, so no need to signup. This helps you to get a user successfully to your platform and then you can take care of your features.
2. Geographical Location of the users : This is required to gauge the user mentality mainly and thus decide the process eg An average american person has a facebook or twitter account, whereas british users still hold onto their Yahoo accounts and hence some changes according to the geography can be done.
3. What information is highly essential before a user joins your system. -
According to me none,check the signup for http://www.jango.com/ , they start with user trying their functionality,
and exposing the remaining stuff when signed up which is simply username and p/w, so lots of stuff happens behind the scenes,
like spam management etc.
The first level of testing we generally do is, asking friends to help but as Abhishek said it not necesarily brings an honest feedback. Hence, we ask people to refer random testers, different users,say for example a 10 yr old boy or a 65 your old person , who are extremes in their nature and log how do they reach to the different parts of the app.
Even before doing a beta,or a sneak peak such a testing is required.
Regards
Saurabh
Best Regards
Saurabh Bhatia
http://safewlabs.com/
http://github.com/saurabhbhatia
Saurabh Bhatia
Finally we had to take a solution similar to rpx,because rpx provides an api and cannot be customised , which is free and open source , JQuery based http://code.google.com/p/openid-selector/ .
We stripped various Open Ids and kept only Google and Yahoo. Others were in the popup if anyone had one of them.
this simple step helped a lot of signups because most of the people had Yahoo or Gmail ids which they used to login.
If you are keeping general users, you can consider, Facebook connect and twitter oauth too.
Regards
Saurabh
AbhishekParolkar
I just found a good app to help in A/B Testing , http://visualwebsiteoptimizer.com/
, thought it might help others too.
Works great for scientifically validating your assumptions on
usability.
- http://github.com/parolkar