I've moved my blog to a new domain. Visit jeffwongdesign.com

Google User Experience

It's nice to see some other perspective at Google other than Marissa Mayer. John is a UX designer at Google. Matt was a co-founder of Feedburner. This is a talk about user experience, not design. There are so many things that affect UX that do not fall in the category of design. These are tips that anyone can use.

Watch the video.


The video's a little long, so here are the notes:

Make your app Fast
Focus on latency. Users should get immediate feedback if they do something. Mozilla optimized their download page by 2.2 seconds and saw their download conversions improved by 15.4%. Speed is important to Google. It takes 470 ms to download the Google search page in Firebug. Firebug lets you see how fast everything downloads, which is useful to analyze your web app.
Other tools are pagetest, page speed and yahoo yslow. Page speed and page test are add-ons to firebug. They will give you advice on how to improve.
Its OK that part of the aesthetic does not load quickly, as long as the content loads quickly.
Fred Wilson, an investor also says that the speed is the most important feature. Users dont care about how hard it is for your app to work, they just care that it works fast.

Make your users fast too
Provide shortcuts - small things add up over time and make the experience frustrating for your users.
Focus on the first field in the form. Saves 1/2 a second, saves a click.
Example: an airline site remembers the last inputs entered by a user.
Big targets are better than small targets when it comes to mouse clicks or mobile devices.
Tab order: users think a certain way when they use your app. Twitter devs made it so that the users go to "remember me" before going to the "sign in" button, even though the sign in button comes first.
Make sure users can submit a form by hitting enter.

Be Yourself
Make sure using the app is fun for the users. Don't have to be all corporate about it.
Don't over explain errors, messages, etc. It only needs to show what the user needs to know.

Engage in conversations with your users
The see you as being on their side and are more likely to forgive you if something goes wrong.
Listen to what people are saying about your product. Monitor blogs, facebook, twitter, etc. to see what people are saying about you. Blog search, technorati, etc. help you do that.
Twitter lets you see stuff that is too new for people to have written a blog post about it, or too small to write an email or a blog post.
When you respond, do it with honesty and humility. Matt has an anecdote here.

Surprise your users
Anecdote: The publisher at Google team decided to build a mobile version of the app. They designed and built it within a week. Someone tweeted about the feature before it was announced, so they called him into the office and showed him the app, which completely surprised him.
Use schwag wisely. It's a great way to build your brand.
Provide your own support: It's a good way to see what's going wrong and to fix it immediately.
Easter eggs that happen when the right things happen is a good way to surprise users.

Be willing to give up control
All of us use 3rd party libraries, use shared code, etc. The same applies to accounting, financing, etc.
For example, whenever feedburner needed anything more than simple graphics, they used istockphoto.com. What if someone else uses the same graphic? That's pretty low risk, and you'll find yourself working with the same artists over time to build your brand.

Be polite
Shoutout to Erica Hall - she gives a great presentation on how to be polite to users.
Be friendly to users, try to understand how your app might annoy users.
There are many error messages that tell you what is wrong (maybe) but no way to fix it or change it.
Make it easy to get started with your app.
Tripit lets you save your travel info. You forward your itinerary to them at an email and they create a shell account that you can sign up with if you want to.
Use Open ID, so people dont have to create a new account unless they have to.
Offer a way out.
Make it easy for your users to leave. No one likes to feel trapped, so don't build dead ends or dark alleys in your app.
Eg. Save changes OR Cancel and Don't save.
E.g. Netflix lets you put your membership on hold and explains what happens if you do.

Prepare for failure
Fast failure makes a lot of sense, but this is not about that.
This is about dead ends and blind alleys. Things that you did not prepare for.
Don't ask people to format phone numbers, enter http:// or ask which kind of card it is. Make messy data entry as easy as possible for the users.
E.g. BBC always has a high and low graphics version of all their content. If you expect a lot of traffic, make sure you have a low-res version of your app that still works, even if the high-res version goes down.
E.g. the "fail whale" at Twitter.

Be reliable
This brings the talk full circle. Speed and reliability are both critical. Stakes are higher for certain apps (banking) and an unreliable site can affect your brand.
Commit as a team to making your service as reliable as possible.

See all the notes here.

No comments:

Post a Comment