The Adventures of Cindy Li

SXSW 2006: Designing the next generation of web apps

March 14 , 2006

Speakers

Williams: Ceo odeo.com, podcasting
Oates:Flickr
Trott:sixapart, blog software
Rodenbeck: Founder, stamen design
Jeffery Veen: www.google.com

Stamen Rodenbeck:

Data visulization for moveon.org
50,000 people online at the same time, real audio.
locate you on the map after you log on and locate people around u.
results of an on-line poll where people would tell the announcers different things. California results of some questions that were offered you could get fine grained to see who was responding and where. Find new applications in an exploratory ways. To fold it back into the client work. Application call

Mappr-to map where the photographs are coming from…

Taking pics of the capitals of teach state. If you start doing tags/searches that aren’t just people and other things you will see route 66..
tagged different things and places, combining it w/the taxonomy.

Cabspotting: the stops the cabs are making and see if they are active. It will go on in ab otu 2 weeks.

IRC: visualization of the back channel (Etech)
Show..taxicab data..
animated over time to see the movement of the cabs. The speed of the taxi cabs. What are you doing w/the visulations. for the exploratory for san fran. Invisible dynamics to see the ebbs and flows of the data.

Nina Trott:

Sixapart
So… iteration, think about product design. What we’ve seen is iteration based on multiple products. Comet is in private beta. An evolution of what we’ve had.. live journal.. social communication such as certain friends are crucial.
Get the next 100million blogging.. they are going to use a tool that is different than what we’re using now. How do you get to that app?
What it takes to get that? APIs? Within the products. IF we need billing systems we want only one. Nothing glamorous. Crucial to get your apps out.
How does splashblog.. acquired Splash. Created 4 mobile clients mobile devices… we’re seeing that to get the next people blogging we’re going to need some people using mobile devices. Posting comments…
back and forth is what its going to take to get the next generation.

Some flickr bits

Now that we know what we’re dealing with (a little more).
We can continue to make it better. When we release a new version we hear it
I’ve got so many people imposing their viw os fwhat flickr is really for that I’ve lost track. Canall you guys who idctate what flickr is..slide changed.. Grrr… Welcoming new people.(important)
If you dont have contacts… showing screen w/o contacts. It explains what you can do.
Its not telling people what they should be doing but what they can do. Slideshow.. people like it.
The organizer is flash based. It is pretyt heavy going. I dont use it currently.  Showing the organize page. We’ve been working on it internally.

Quote form users: “I can’t see it in that group and i need to be informed, cause flickr is a giant chaos and now I was having the control of it!!!!

Evan Williams: we work at odeo on audio stuff. Bring audio to the web and listen to audio based on audio tools. People like to see things when they look on the screeen: audience laughter:::
(Demo of the odeo studio.)
Where people can see a message and listen it?  They can leave a web message.  Recorder interface that is implemented.

Veen: Light weight interface..?

Williams: Limitations on how sophisticated of a tool on a web browser. This tiny verison is more popular than the complicated version. 3 min limit in both versions. Director of podcasts that are short. We’re trying to suggest is a quick little voice message is a lot less intimidating.
Veen: gateway drug?
Williams: laughs…

Veen: taking those desires and trying to solve the problems that were solved.
Simple enough to make peopl ecare about web analytics. There is very little financial incentive.

http://www.measuremap.com unbelieveable power, server logs.

How are people participating, what effect do I have on the outside world?

How are they getting in through links in?
They didn’t care what flash or ajax is.
People really want to work like magic.
Iteration: I developed measuer map from .. as a designer not a development focus orgazniation. We started by designing it and then figured out how to build it. We ignored the technology.
Iteration we had a bunch at one time.

The way to maximize user experience is…

unquestionably through interation. David Hornik, ventureBlog

We’ve probably had more releases already than Word did in its first decade. Sam Schillace, Writely

Trott: Its been very different for us as a process when we first started. My hubby Ben and I before. We could sit there and know what each other’s strenghs was and what each other was doing. It was fun but we were very spoiled. We have to communicate in a team.  it doesn’t get smaller than a wife and hubby in a bedroom. :) Two people in a room working and you introduce more people. 125 people.. designers/engineer….
a lot is based on we have 4 products. There aren’t 30 people working on livejournal. The iteration comes from learning what we’re doing based on a product. Nothign is easy anymore. Figuring out what works. You ahve to get the product out. Get people who are working on things testing it to work out the bugs. Movabletyle and live journal. Huge support of commercial clients. Users don’t want things rushed. LiveJournal isn’t polished but it works…

Get your product out

Veen: measure map.. we were releasing as quick as we could. twice a week.
monday/thursday gave us a periodic cycle. see how far we could and then move on. The cycle was better for us

Oates: Dev environment to check and click deploy. Any of the dev could deploy anytime they watned. People in the early days could change and deploy it. See it live.

Veen: Frustration was my lack of ability to do frequent iteration. The startups we worked with were more amiable toward the release cycle.

Rodenbeck: reiteration chart:http://stamen.com/studio/solvingproblems

Oates: The solutions you try may help your problem. The instant feedback was valuable

Williams: We adopted a scrub. We started w/30 day iteration we cut it down to 2weeks. Scrub is a methodology. Releasable in 30days… engineers said it was taking too long. In some cases we keep the release cycle. In the two weeks and 30 days we decide what we’re going to work on. Its not thinking of a new feature and derailing.

One of the things that people are going to want to do….
how do you figure it out. What we saw was one of the bloggers were more happy writing to their friends/family.. or no one. Seeing what tehy were doing and see what the activity level was.

Williams: people like getting the voicemails. If you record something default its private. It looks the same as the public version.. eh…
it made people more willing to do it. There is a gray area that it is private to the extent you want it to be private.

Veen:Editorial control is coming from algorithms now: features coming out of people using my tools whenever you want.

Veen flickr…

sort by most interesting… what is that?
It was written as an algorithm..
u dont know what comes in..

trott: Linkedin was not going to have photos in profiles.
keeping it to finding a job.
we never wanted emoticons. we said no. we watned more of css emoticons didn’t make sense for us. we make these decisions and ti amkes a difference in presentations.
veen: design w/o stuff…
todenbeck: we started working with digg w/their data and we haven’t made a comp, we’re poking through the data right now. teh editorializing will come out of the data, from the top.

User generated content means design becomes a container

takign google maps and the crime data.. so you know waht is going on in your neighborhood
chicagocrime.org

trott: advice as a designer.. when i do a comp. i wil not put greek text..
put in real data. The minute you let someone in it is containers. You are in an application but you are in art. I care more about the content is presented easily. I would rather my stuff easily communicated.

Oates: showing what is possible…for the users

Rodenbeck: its ok if you just want coffee… design is a container. its a good thing :) It’s cool. Adtive and you design for something that is life. 

People get afraid of the template…

Williams: w/u actually designing containers for people to create conatiners. People can totally screw w/the design.

Feedburner
Right upfront they give you detail on how you can quit. Services they offer to help you get out of it. There is no insecurity.
They are going to be competitive by allowing you out.

How do you filter the crazy people?

oates: Difficult voice is the loudest.
veen: lycos.. their customer support 19 people wrote they didn’t like that blue.. feedback is constantly coming in. ligthening some out of hand.
use that to synthesize something amazing.

trott: we get updates of what people are asking. they are in the trenches.
they are distilling what exactly is working and whwat isn’t. what’s should be a feature.

willliams.. we roll back when its broken.
oates:you want to think its good and creative

question:How do you balance the customer requests over what people dont know what they want..?

Trott: innovation of the features… if we only did what people requested then we weren’t evolving.
Veen: use to innovate.. foundation from which you write good content.
Book: Abstracting craft.. malcom, the way digital tools are becoming a medium. As the tools are better they are more fine grained. Put your ahnd in here and shape the clay.