it20070312 - Monday

Panel:

• Cheng
• James
• Poisson (radar)
• Wilhelm
• King

(They changed rooms without telling anyone. I was 10 minutes late, but at least I had company.)

Jones: Does some prototyping by putting post-it notes on blocks of wood.

King: User testing for mobile (while person is texting while driving or whatever) is more difficult than watching someone us a PC.

Wilhelm: Have people walk around with a block and ask them what they’d like to do with that block.

Q: What kind of metrics do you use when designing for mobile?

Poisson: Clicks are gold. Saving an extra click, multiplied millions of customers, is a lot of effort saved. Used radar.net instead of .com because .net is easier to type on a mobile device.

King: Does a lot with click tracking as well.

Wilhelm: Assume a lot of upload failure and interruption when dealing with files. Make sure that kind of issue handling is built in from the start.

Poisson: You need to think about about time limits and attention span limits much more than you would on a desktop application.

Jones: Need to think of a three-second attention span. No clicking off rules and conditions screens.

Q: You have all these inputs and outputs… Text, voice, photos, etc… What makes you try to write a custom app rather than shove everything over SMS?

King: Depends on what you need to do? Sending photos over SMS would suck.

Q; At Nokia, how much time do you spend designing beyond the phone?

Jones: It’s more about about the interraction between the cloud and the little thing you’re communicating with. More focus on connections to web services. Mobile makers kind of make the equivalent of a wiki stub asking themselves or others to fill it in later. By doing that pattern of stub-making gets you to open really interesting ideas.

Q: Flash lite is not supported by majority of phones. How does one go about testing one’s app on all these phones when emulators are not that great?

King: Flash Lite is an excellent prototyping tool, but not ready for actual design yet.

Poisson: For applications, you can’t build once and use everywhere. You need to know going in you need to prep for a range of devices. Know the platforms your users use and try to build for them.

Wilhelm: deviceanywhere, etc. You can check out and test other phones online. Not 100% accurate, but better than buying every different model.

Jones: Look at the presentation by Brian Fling yesterday. (My notes from Fling’s presentation.) Get five or so varried phones, and test against them.

Wilhelm: Try to group devices and find similarities and then test against examples in those categories.

Poisson: Join developer networks from mobile manufacturers. Lots of good information from them.

Q: What are your practical strategies on getting people to use your applications? Off-deck or on-deck?

Wilhelm: Off-deck: Distribution not managed by carrier. On-deck: Distribution by carrier. Off deck allows full-control of experience for user. Lots of pain to get on-deck. Some carriers have a lot of rules and regulations about what you can do in your application.

Poisson: Carriers still have some control when you’re off-deck. Just get your app out there, see what people want to do and go from there. If it works, carriers may come to you about tying your tool to their device.

Wilhelm: Maybe build teirs into your service.

Q: What do you think of development of browsers for phones?

King: Opera browser for mobile is great.

Jones: Look at what Fling said. Pare site down to attention-friendly components.

Q: Do you have any favorite resource for mobile design patterns?

Jones: W3C Mobile Web Initiative… Plus developer networks. We’re still learning this stuff as we go. Look up small services blog.

Panel:

• Kristian Bengtsson, FutureLab
• Jeffrey Zeldman, Happy Cog, A List Apart
• Chris Messina, Citizen Agency, co-founder BarCamp
• Luke Wroblewski, Yahoo! Social Media
• Liz Danzico, Bobulate

Danzico: What does unstuck mean? How can we take risks and make changes, convince clients to do the same, break old processes that we rely on? Avoid the same thing over and over.

(Traffic slide: New Yorker: Looks like good traffic. Pennsylvanian: Looks like a traffic jam.)

Unstuck: The act or process of doing good work, being productive, feeling fulfilled on a team.

Q: Let’s hear about your personal processes…

Messina: Finds that areas he gets stuck are where he keeps his ideas to himself.

Zeldman: Management through conversation. Talk quite a bit, build trust and comfort, before showing prototypes, etc. No process or rules, just tries to listen.

Wroblewski: Ongoing client feedback loop. Always try to be engaged with information flowing in. Staying on top of trends online and putting that out to the customer.

Bengtsson: Be fearless and have fun. “It’s better to be a flamboyant failure than a mediocre success.” - Sex pistols singer.

Wroblewski: Look into book the “Five Dysfunctions of a Team: A Leadership Fable.” Absence of trust is at the base of it.

Zeldman: Remember we live in a soundbyte culture. If you can think of a soundbyte, use it over and over again, you can convince people.

Bengtsson: Thousands of ideas in your head. Write them down, sketch them down, you’ll realize that most suck. One or two may still seem good enough to work with, and you can throw those out to people to see what they think.

Q: Is design a tool to get you unstuck?

Bengtsson: Design is both the process and the goal. There shouldn’t be a gap between design and development. Start on agreement on the design throughout an organization.

Wroblewski: Don’t say you need to get the design team involved. You’ve already set up a wall by separating groups.

Messina: Building software is a political process, so you need to understand everyone’s needs and interests and mesh that into your overall process. Don’t have three wikis–one for design, development and sales–have one huge wiki everyone can see and contribute to. Don’t wikify in little pieces. Let communication flow everywhere. One team, not multiple teams.

Zeldman: Subdivide into small teams if possible. Make sure contributors who don’t know all the terms or technical issues still feel loved and listened to. Get buy in and trust before making prototypes.

Wroblewski: Small teams and big teams don’t necessarily have different dynamics. What’s true for small teams is true for big teams. You just need more love and openess.

Lightning Round!

Q: Stuck with own software and hosting — constantly have slap on new coats of paint, etc. How do we move forward?

Zeldman: Quit.

Q: Clients focus on what they want whether than their clients need. How to refocus them?

Zeldman: Say no to bad jobs until you get the right one.

Messina: Get users to use tool in front of client and show client they don’t know what they’re talking about it.

Wroblewski: Make two columns, user goals, business needs. Clients need to acknowledge user needs… hopefully.

Q: How do I get others to embrace standards?

Wroblewski: Show them how it will benefit them, what it means for them.

Zeldman: Wrote his designing with web standards book for your boss, not for developers.

Messina: Come to a story that can be repeated back to you.

Zeldman: In other session yesterday, blind person w/ lukemia trying to get info from medical websites… and unable to read it because poor use of standards and lack of accessibility.

Q: Multi-million project that most agree is doomed to failure, management won’t kill it because of amount of money spent. How to get people to kill it?

Zeldman: Use it as a demo of what doesn’t work.

Wroblewski: If you’ve spent that much money there has to be something good in it. Can you find a way to save the project?

Messina: Fail quickly and reassess often.

Q: Everyone on a team wants to make decisions. How do you get a decision to stick.

Zeldman: What does the boss say? He may need to step up and make the decision.